SAMPLING AND ANALYSIS
PROCEDURES FOR
SCREENING OF
INDUSTRIAL EFFLUENTS
FOR PRIORITY POLLUTANTS
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
EFFLUENT GUIDELINES DIVISION
WASHINGTON, D.C.
MARCH 1977 revised APRIL 1977
-------
FOREWORD
These guidelines for sampling and analysis of industrial wastes
have been prepared by the staff of the Environmental Monitoring and
Support Laboratory, at the request of the Effluent Guidelines Division,
•
Office of Water and Hazardous Wastes, and with the cooperation of the
Environmental Research Laboratory, Athens, Georgia. The procedures
represent the current state-of-the-art but improvements are anticipated
as more experience with a wide variety of industrial wastes is obtained.
Users of these methods are encouraged to identify problems encountered and
assist in updating the test procedures by contacting the Environmental
Monitoring and Support Laboratory, EPA, Cincinnati, Ohio 45258.
-------
CONTENTS
ORGANICS BY PURGE AND TRAP - GAS CHROMATOGRAPHY 1
Scope 1
Special Apparatus and Materials . . . , 2
Gas Chromatographic Column Materials 2
Procedure 3
X"
Preparation of Standards 3
Preliminary Treatment of Sample 4
«
Purging and Trapping Procedure 4
GC-MS Determination 6
Purge Parameters 6
Gas Chromatographic Parameters 7
Mass Spectrometer Parameters 7
Quality Assurance 8
Precision 9
Calibration of GC-MS System 10
Qualitative and Quantitative Determination .... 10
Reporting of Data 11
Direct Aqueous Injection Gas Chromatography 11
ORGANICS BY LIQUID-LIQUID EXTRACTION - GAS CHROMATOGRAPHY . . 16
Scope 16
Special Apparatus and Materials 16
Procedure 16
Base-Neutral Extraction 17
-------
Separatory Funnel Extraction 17
Acid (Phenols) Extraction 18
Emulsions 19
Continuous Extraction ....... 20
Blank Extraction 20
s'
Pesticides 21
GC-MS Analysis 23
•
Base-Neutral 23
Acid 25
Quality Assurance 29
Reporting of Data 30
METALS 43
Sample Preparation ..... 43
Apparatus 44
Procedure 44
Quality Assurance ... 47
Data Reporting 4,8
CYANIDES 49
»
Sample Preparation 49
t
Sample Procedure 49
Quality Assurance 49
Reporting of Data 49
-------
PHENOLS 50
Sample Preparation 50
Procedure 50
Quality Assurance .... 50
Reporting of Data 50
REFERENCES 51
APPENDIX I 53
APPENDIX II 55
APPENDIX III 63
APPENDIX IV 70
-------
LIST OF TABLES
Table I. Elution Order of Volatile Priority Pollutants ... 12
Table II. Characteristic Ions of Volatile Organics 14
Table III. Pesticides 33
Table IV. Base-Neutral Extractables 34
Table V. Acid Extractables 36
Table VI. Elution Order of Most of the Semivolatile
Priority Pollutants 37
Table VII. Order of Elution for OV-17 SCOT Column 41
Table VIII. Metals 45
-------
Appendices
I. General Information
II. Possible Sources for Some Priority Pollutant Standards
III. Collection of Samples for Screening Analyses
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
SUBJECT: Sampling and Analysis Procedures for Screening
of Industrial Effluents for Priority Pollutants
FROM: William A. Telliard, Chief
Energy and Mining Branch
TO: Project Officers
Effluent Guidelines Division
THRU: Robert B. Schaffer, Director,
Effluent Guidelines Divisi
DATE:
MAY 2 7 1977
As you know, in the s-ettlement of several cases in the District
Court for the District of Columbia, the Environmental Protection
Agency has agreed to review and revise regulations based on the
Best Available Technology Economically Achievable (BAT), New
Source Performance Standards, and Pretreatment Standards for 21
industrial categories.
In this revision, consideration is to be given to the application
of limitations of a list of 65 materials appearing in Appendix A
of the Settlement Agreement. These materials are generally
referred to as priority pollutants. The priority pollutants are
both single compounds and families of compounds. The Agency has
established an unambiguous list of 129 compounds which it
believes fulfills the requirements of the court order and can be
analytically determined.
To maintain consistent sampling and analytical procedures the
Agency has developed a sampling protocol and analytical methods
to be used for screening for priority pollutants. This protocol
represents the most current procedures for the sampling and
analysis of these priority pollutants. Because of the large
number of analysis required, argon plasma atomic-emission
spectroscopy will be used by the Agency for most metals analysis,
Pertinent information about this analytical method, which is an
accepted alternate method under section 304(g) is attached.
The data gathering process basic to revising the regulations
consists of two phases. The initial phase is the screening
sampling and analysis procedure to ascertain the presence or •
EPA Form 1320-6 (Rev. 6-72)
-------
absence of the priority pollutants. The second phase or
verification sampling will be used to quantify those pollutants
found to be present during the screen sampling.
These materials are made available for your information and use
during the screening phase only.
-------
UNITED t,. ATES ENVIRONMENTAL PROTECTIL.. AGENCY
SUBJECT: Approval of Alternate Test Procedure - Automated DATE: . . »• O 9 19/6
Simultaneous Analysis of Twenty Selected Elements
by Inductively Coupled Argon Plasma Emission Spectrotcopy
FROM: Mr. Francis T. Mayo
Regional Administrator, Region V
TO: Mr. Thomas E. Yeates, Director
Central Regional Laboratory, Region V
THRU: Mr. Chris Timm, Director ' /
Surveillance & Analysis Division, Region ^/
The Environmental Monitoring and Support Laboratory (EMSL)-- Cincinnati,
EPA has carefully reviewed your application for use of an alternate test
procedure for the automated simultaneous analysis of twenty elements by
emission spectroscopy using the inductively coupled argon plasma as the
emission source. Your application specifies that the automated instrumen-
tal methodology will be used at the Central Regional Laboratory, Region V
for all sample types applicable to the National Pollutant Discharge Elimi-
nation System (NPDES).
>
The proposed method uses the sample digestion procedure of 40 CFR Part 136,
but instead of utilizing referee atomic absorption spectrometry, the
digested sample is aspirated into a high temperature inductively coupled
argon plasma (ICAP), and several total elemental concentrations in the
aspirated sample are measured simultaneously using an emission spectro-
graph with an appropriate photomultiplier tube for each element. The
calculation of the elemental concentrations is done by a computer inter-
faced to the spectrograph. Twenty elements, primarily metals, have been
selected by the Central Regional Laboratory for analysis by ICAP-emission
spectroscopy. In addition to total elemental concentrations, the proposed
methodology can readily measure dissolved concentrations for the same
elements simply by filtering a suitable aliquot through a 0.45 y membrane,
acidifying the filtrate as necessary for preservation, eliminating the
digestion procedure and aspirating the filtrate into the ICAP-emission
spectrograph.
The comparability data you have provided for ICAP-emission spectroscopy
and referee atomic absorption spectrophotometry indicate the two methods
yield equivalent data for a variety of waste effluents representative of
the NPDES. ICAP-emission spectroscopy is shown to provide a comparable
or superior performance, depending on the element, for the measurement of
recovery and precision for random element "spikes" of NPDES waste effluents.
Although the stated detection lira-its and lowest quantitatively deterroinable
concentration vary slightly from day-to-day and are a function of the ICAP
nebulizer, the reportable detection limits for ICAP-emission spectroscopy
EPA PM» 1320-4 (ft.*.
-------
Organics by Purge and Trap -
Gas Chromatography
1. Scope
This method is designed to determine those "unambiguous
priority pollutants," associated with the Consent Decree, that
are amenable to the purge and trap method . These compounds
are listed in Table I of this section. It is a gas chromato-
graphic-mass spectrometric (GC-MS) method intended for quali-
tative and semi-quantitative determination of these compounds
during the survey phase of the industrial effluent study.
Certain compounds, acrolein and acrylonitrile, are not
efficiently recovered by this method and should be determined
by direct aqueous injection GC-MS. Direct aqueous injection
GC-MS is recommended for all compounds that exceed 1000 ug/1.
The purge and trap and the liquid-liquid extraction methods
are complementary to one another. There is an area of overlap
between the two and some compounds may be recovered by either
method. However, the efficiency of recovery depends on the
vapor pressure and water solubility of the compounds involved.
Generally, the area of overlap may be identified by compounds
boiling between 130°C and 150°C with a water solubility of
approximately two percent. When compounds are efficiently re-
covered by both methods, the chromatography determined the
method of choice. The gas chromatographic conditions selected
-------
- 2 -
for the purge and trap method are, generally, not suitable for
the determination of compounds eluting later than chlorobenzene
2. Special Apparatus and Materials
Sample extraction apparatus (minimum requirements):
5-ml glass syringes with Luer-Lok - 3 each
2-way syringe valves (Teflon or Kel-F) - 3 each
8-inch, 20 gauge syringe needle - 2 each
5-ml glass, gas-tight syringe, pressure-lok
or equivalent - 1 each
Tekmar Liquid Sample Concentrator, model LSC-1
or equivalent. Includes a sorbent trap
consisting of 1/8 in. O.D. (0.09 to 0.105
in.I.D.) x6in. long stainless steel tube
packed with 4 inches of Tenax-GC (60/80 mesh
and 2 inches of Davison Type-15 silica gel
(35/60 mesh).
3. Gas Chromatographic Column Materials
Stainless steel tubing 1/8 in. O.D. (0.09 to 0.105 in.
I.D.) by 8 ft. long. Carbopack C (60/80 mesh) coated with
(c)
0.2% Carbowax 1500 . Chromosorb-W (60-80 mesh) coated with
3% Carbowax 1500.
(a) Available from Precision Sampling Corp., P.O. Box 15119,
Baton Rouge, LA 70815.
(b) Available from Tekmar Company, P.O. Box 37202,
Cincinnati, OH 45222.
(c) Available from Supelco, Supelco Park, Beliefonte, PA
16823. Stock No. 1-1826.
-------
- 3 -
4. Procedure
Preparation of Standards - Prepare standard stock solu-
tions (approximately 2 ug/yl) by adding, from a 100 yl'syringe,
1 to 2 drops of the 99+% pure reference standard to methanol
(9.8 ml) contained in a tared 10 ml volumetric flask (weighed
to nearest 0.1 mg). Add the compound so that the two drops
fall into the alcohol and do not contact the neck of the flask.
Use the weight gain to calculate the concentration of the stand-
ard. Prepare gaseous standards, i.e., vinyl chloride, in a
similar manner using a 5 ml valved gas-tight syringe with a
2 in. needle. Fill the syringe (5.0 ml) with the gaseous com-
pound. Weigh the 10 ml volumetric flask containing 9.8 ml of
methyl alcohol to 0.1 mg. Lower the syringe needle to about
5 mm above the methyl alcohol meniscus. Slowly inject the
standard into the flask. The gas rapidly dissolves in the
methyl alcohol. Reweigh the flask, dilute to volume, mix,
tightly stopper, and store in a freezer. Such standards are
generally stable for at least one week when maintained, at less
than 0°C. Stack standards of compounds which boil above room
temperature are generally stable for at least four weeks when
stored at 4°C.
[Safety Caution: Because of the toxicity of most' organo-
halides, primary dilutions must be prepared in a hood. Fur-
ther, it is advisable to use an approved respirator when
handling high concentration of such materials.]
-------
- 4 -
From the primary dilution prepare a secondary dilution
mixture in methyl alcohol so that 20.0 yl of the standard,
diluted to 100.0 ml in organic free water, will give a stand-
ard which produces a response close to that of the unknown.
Also prepare a complex test mixture at a concentration of
100 ng/yl containing each of the compounds to be determined.
Prepare a 20 yg/1 quality check sample from the 100 ng/ul
standard by dosing 20.0 ul into 100.0 ml of organic free water,
Internal Standard Dosing Solution - From stock standard
solutions prepared as above, add a volume to give 1000 yg each
of bromochloromethane, 2-bromo-l-chloropropane, and 1,4-
dichlorobutane to 45 ml of organic free (blank water) con-
tained in a 50 ml volumetric flask, mix and dilute to volume.
Prepare a fresh internal standard on a weekly basis. Dose
the internal standard mixture into every sample and reference
standard analyzed.
Preliminary Treatment of Sample - Remove samples from
cold storage (approximately an hour prior to analysis) and
bring to room temperature by placing in a warm water bath
at 20-25°C.
Purging and Trapping Procedure - Adjust the helium purge
gas flow to 40 ml/min. Set the Tekmar 2-way valve to the
purge position and open the purging device inlet. Remove
the plungers from two 5-ml syringes and attach a closed 2-way
syringe valve to each. Open the sample bottle and carefully
-------
- 5 -
pour the sample into one of the syringes until it overflows.
Replace the syringe plunger and compress the sample. Open
the syringe valve and vent any residual air while carefully
adjusting the volume to 5.0 ml. Then close the valve. Fill
the second syringe in an identical manner from the same
sample bottle. Use the second syringe for a duplicate analy-
sis as needed. Open the syringe valve and introduce 5.0 yl of
the internal standard mixture through the valve bore, then
close the valve. Attach the 8-inch needle to the syringe
valve and inject the sample into the purging device. Seal
the purging device and purge the sample for 12 minutes. The
purged organics are sorbed on the Tenax-silica gel trap at
room temperature (20-25°C).
While the sample is being purged, cool the gas chromato-
graphic column oven to near room temperature (20-30°C). To
do this, turn heater off and open column oven door.
At,the completion of the 12-minute purge time, inject
the sample into the gas chromatograph by turning the valve
to the desorb position. Hold in this position for four min-
utes while rapidly heating the trap oven to 180QC, then return
the valve to the purge position, close the GC column oven
door, and rapidly heat the GC oven to 60°C. Consider this
time zero and begin to collect retention data. Hold at 60°C
for four minutes, then program at 8°/minute to 170°C and hold
until all compounds have eluted. Begin collecting GC-MS
-------
- 6 -
GC-MS data as soon as the GC-MS vacuum system has stabilized
(<10~5 torr).
While the sample is being chromatographed, flush the
purging device with two 5-ml volumes of organic free water.
Then bake out the trap (vent to atmosphere) to minimize the
amount of water desorbed into the GC-MS system during the
succeeding injection step. [Note: If this bake out step
is omitted, the amount of water entering the GC-MS system
will progressively increase causing deterioration of and
potential shut down of the system.]
GC-MS Determination - Suggested analytical conditions
for determination of the priority pollutants amenable to
purge and trap, using the Tekmar LSC-1 and the computerized
Finnigan 1015 GC-MS are given below. Operating conditions
vary from one system to another; therefore, each analyst
must optimize the conditions for his equipment.
Purge Parameters
Purge gas - Helium, high purity grade
Purge time - 12 minutes
Purge flow - 40 ml/min.
Trap dimensions - 1/8 in. O.D. (0.09.to 0.105 in. I.D.)
x 6 in. long
Trap sorbent - Tenax-GC, 60/80 mesh (4 in.) plus Type 15
silica gel, 35/60 mesh (2 in.)
Desorption flow - 20 ml/min.
Desorption time - 4 min.
Desorption temperature - 180°C
-------
- 7 -
Gas Chromatographic Parameters
Column - Stainless steel, 8 ft. long x 1/8 in. O.D.
(0.09 to 0.105 in. I.D.) packed with Carbopack C
(60/80 mesh) coated with 0.2% Carbowax 1500, pre-
ceded by a 1 ft. x 1/8 in. O.D. (0.09 to 0.105 in.
I.D.) packed with Chromosorb-W coated with 3%
Carbowax 1500.
Carrier gas - Helium at 33 ml/min.
Oven temperature.- Room temperature during trap desorp-
tion, then rapidly heat to 60°C, hold at 60°C for four min-
utes, then program to 170°C at 8°/minute. Hold at 170°C for
12 minutes or until all compounds have eluted.
Mass Spectrometer Parameters
Data system - System industries System 150
Separator - glass jet
Electron energy - 70 ev
Emission current - 500 ua
Ion energy - 6 volts
Lens voltage - (-)IOO volts
Extractor voltage - 8 volts
Mass range - 20-27, 33-260 amu
Integration time/amu - 17 milliseconds
Samples/amu - 1
Gas Chromatographic Column Conditioning Procedure; -
Attach the Carbowax 1500-Chromosorb end of the column to the
inlet system of the gas chromatograph. Do not, at this time,
-------
- 8 -
attach the column exit to the detector. Adjust the helium
flow rate through the column to 33 ml/minute. Allow the
column to flush with helium for ten minutes at room tempera-
ture, then program the oven from room temperature to 190°C at
4°C/rainute. Maintain the oven at 190°C overnight (16 hours).
Handle the column with extreme care once it has been
conditioned because the Carbopack is fragile and easily frac-
tured. Once fractured, active sites are exposed resulting in
poor peak geometry (loss of theoretical plates). Recondition-
ing, generally, revitalizes the analytical column. Once
properly conditioned, the precolumn may be removed. The re-
tention data1listed in Table I was collected with the pre-
column in the system.
Quality Assurance - The analysis of blanks is most
important in the purge and trap technique since the purging
device and the trap can be contaminated by residues from
very concentrated samples or by vapors in the laboratory. Pre-
pare blanks by filling a sample bottle with low-organic water
(blank water) that has been prepared by passing distilled
water through a pretested activated carbon column. Blanks
should be sealed, stored at 4°C, and analyzed with each group
of samples.
After each sample analysis, thoroughly, flush the purg-
ing device with blank water and bake out the system. Sub-
sequently, analyze a sample blank (one that has been transported
-------
- 9 -
to and from the sampling site) . If positive interferences
are noted, analyze a fresh laboratory sample of blank water.
If positive interference still occurs, repeat the laboratory
blank analysis. If interference persists, dismantle the
system, thoroughly, clean all parts that the sample, purge
gas and carrier gas comes into contact with and replace or
repack the sorbent trap and change purge and carrier gas.
Precision - Determine the precision of the method by
dosing blank .water with the compounds selected as internal
standards - bromochloromethane, 2-bromo-l-chloropropane, and
1,4-dichlorobutane - and running replicate analyses. These
compounds represent early, middle, and late eluters over the
range of the Consent Decree compounds and are not, themselves,
included on the list. Construct Quality Control charts from
the data obtained according to directions in Reference 9.
The sample matrix can affect the purging efficiencies
of individual compounds; therefore, each sample must be
dosed with the internal standards and analyzed in a manner
identical to the internal standards in blank water. When
the results of the dosed sample analyses show a deviation
greater than two sigma, repeat the dosed sample analyses.
If the deviation is again greater than two sigma, dose
another aliquot of the same sample with the compounds of
interest at approximately two times the measured values and
analyze. Calculate the recovery for the individual compounds
using these data.*
*See Reporting of Data Section, p. 11.
-------
- 10 -
Calibration of the gas chromatography-mass spectrometry
(GC-MS) system - Evaluate the system performance each day
that it is to be used for the analysis of samples or blanks.
Inject a sample of 20 nanograms of decalfuorotriphenyl-
phosphine and plot the mass spectrum. The criteria in
Reference 2 must be met and all plots from the performance
evaluation, documented and retained as proof of valid
performance. " ?
Analyze the 20 yg/1 standard to demonstrate instrument
performance for these compounds.
Qualitative and Quantitative Determination - The char-
acteristic masses or mass ranges listed in Table II of this
section are used for qualitative and quantitative determination
of volatile priority pollutants. They are used to obtain an
(e)
extracted ion current profile (EICP) for each compound.
For very low concentrations, the same masses may be used for
selected ion monitoring (SIM) . The primary ions to be used
to quantify each compound are also listed. If the sample pro-
duces an interference for the primary ion, use a secondary
ion to quantify-
(d) Available from PCR, Inc., Gainesville, FL.
(e) EICP is the reduction of mass spectrometric data
acquired by continuous, repetitive measurement of
spectra by plotting the change in relative abundance
of one or several ions as a function of time.
(f) SIM is the use of a mass spectrometer as a substance
selective detector by measuring the mass spectrometric
response at one or several characteristic masses in
real time.
-------
- 11 -
Quantify samples by comparing the area of a single mass
(see Table II) of the unknown in a sample to that of a stan-
dard. When positive responses are observed, prepare and
analyze a reference standard so that the standard response
closely approximates the sample response. Calculate the con-
centration in the sample as follows:
(Area for unknown)
(Area for standard) _ ,, - ,,„,,„--„,
7t r—rr~- =— j. j j i /i \ U9/1 °f unknown
Concentration or standard (vig/1) 3
5. Reporting of Data
Report all results to two significant figures or to the
nearest 10 yg/1. Report internal standard data to two signif-
icant figures.
As the analyses are completed, transfer GC-MS data to
magnetic tape .as described under reporting of data in method
for "Organics by Liquid-Liquid Extraction - Gas Chromatography."
Report all quality control (QC) data along with the
analytical results for the samples. In addition, forward
all QC data to EMSL, Cincinnati.
6. Direct Aqueous Injection Gas Chromatography
As noted in the Scope, Acrolein and acrylonitrile should
be analyzed by direct aqueous injection gas chromatography-
mass spectrometry. See references (3) , (4) -, and (5) for
these methods. The detection level for these methods is 0.1
mg/1 and above.
-------
- 12 -
Table I
Elution Order of Volatile Priority Pollutants
(a)
Compound
chlorome thane"
dichlorodifluoromethane
bromomethane
vinyl chloride
chloroethane
methylene chloride
trichlorofluoromethane
1,1-dichloroethylene
bromochloromethane(IS)
1,1-dichloroethane
trans-1,2-dichloroethylene
chloroform
1,2-dichloroethane
1,1,1-trichloroethane
carbon tetrachloride
bromodichloromethane
bis-chloromethyl ether
1,2-dichloropropane
trans-1,3-dichloropropene
trichloroethylene
dibromochloromethane
cis-1,3-dichloropropene
1,1,2-trichloroethane
benzene
2-chloroethylvinyl ether
2-bromo-l-chloropropane(IS)
bromoform
1,1,2,2-tetrachloroethene
1,1,2,2-tetrachloroethane
(b)
Purging
Efficiency
(percent)
Purging
Efficiency
Modified
Method
(percent)
0.152
0.172
0.181
0.186
0.204
0.292
0.372
0.380
0.457
0.469
0.493
0.557
0.600
0.672
0.684
0.750
0.760
0.818
0.847
0.867
0.931
0.913
0.913
0.937
0.992
1.000
1.115
1.262
1.281
91
0 100 (c)
85
101
90
76
96
97
88
89
92
95
98
94
87
92
0
92
90
89
87
85
88
no data
no data
92
71
88
58
-------
- 13 -
Table I (cont'd)
Compound
1,4-dichlorobutane(IS)
toluene
chlorobenzene
ethylbenzene
acrolein
acrylonitrile
RRT
(b)
1.312
1.341
1.489
1.814
unknown
unknown
Purging
Efficiency
(percent)
74
no data
89
no data
12
no data
Purging
Efficiency
Modified
Method
(percent)
74
(e)
(a) These data were obtained under the following conditions
GC column - stainless steel, 8 ft. long x 0.1 in. I.D.
packed with Carbopack C (60/80 mesh), coated with 0.2%
Carbowax 1500; preceeded by a 1 ft. long x 0.1 in. I.D.
column packed with Chromosorb W coated with 3% Carbowax
1500; carrier flow - 40 ml/min.; oven temperature -
initial 60°C held for 3 min., programmed 8°C/min. to
160°C and held until all compounds eluted. The purge
and trap system used was constructed by EPA. Under
optimized conditions, commercial systems will provide
equivalent results.
(b) Retention times relative to 2-bromo-l-chloropropane
with an absolute retention time of 829 seconds.
(c) No measurable recovery using standard purging and trap-
ping conditions. Under modified conditions, i.e.,
purging at 10 ml/min. for 12 min., recovery is 100%.
(d) Bis-chloromethyl ether has a very short half-life in
water and is not likely to be detected in water.
(e) Recovery 12% under standard purging conditions, i.e.,
room temperature, 30% at 55°C, and 74% at 95°C.
-------
- 14 -
l'
Table II
Characteristic Ions of Volatile Organics
Compound
chloromethane
dichlorodi fluoromethane
bromomethane
vinyl chloride
chloroethane
methylene chloride
trichlorofluoromethane
1,1-dichloroethylene
bromochloromethane(IS)
1,1-dichloroethane
trans-1,2-dichloroethylene
chloroform
1,2-dichloroethane
1,1,1-trichloroethane
carbon tetrachloride
bromodichloromethane
bis-chloromethyl ether
• 1,2-dichloropropane
trans-1,3-dichloropropene
trichloroethylene
dibromochloromethane
cis-1,3-dichloropropene
El Ions (Relative
intensity)
Ion used to
quantify
50(100) ; 52(33)
85(100) ; 87(33) ;
101(13); 103(9)
94(100) ; 96(94)
62(100) ; 64(33)
64(100) -, 66(33)
49(100>;51(33) ;
84(86}; 86(55)
101(100) ; 103(66)
61(100) ; 96(80) ; 98(53)
50
101
94
62
64
84
101
96
49(100); 130(88);
128(70); 51(33)
63(100); 65(33); 83(13);
85(8); 98(7); 100(4)
61(100)
83(100)
96(90)
85(66)
98(57)
62(100); 64(33);
98(23) ; 100(15)
98(100); 99(66);
117(17); 119(16)
117(100); 119(96); 121(30)
83(100)
127(13)
85(66);
129(17)
79(100); 81(33)
63(100)
112(4);
65(33);
114(3)
75(100): 77(33)
95(100)
130(90)
97(66);
132(85)
129(100) ; 127(78)
208(13) ; 206(10)
75(100); 77(33)
128
63
96
83
98
97
117
127
79
112
75
130
127
75
-------
- 15 -
Compound
1f1,2-trichloroethane
benzene
2-chloroethylvinyl ether
2-bromo-l-chloropropane(IS)
bromoform
1,1,2,2-tetrachloroethene
1,1,2,2-tetrachloroethane
1,4-dichlorobutane(IS)
toluene
chlorobenzene
ethylbenzene
acrolein
acrylonitrile
Table (cont'd)
El Ions (Relative
intensity)
83(95); 85(60); 97(100);
99(63); 132(9); 134(8)
78(100)
63(95); 65(32); 106(18)
77(100); 79(33);156(5)
171(50) ,-173(100) ; 175(50)
250(4); 252(11); 254(11);
256(4)
129(64)
164(78)
131(62);
166(100)
83(100); 85(66); 131(7)
133(7); 166(5); 168(6)
55(100); 90(30); 92(10)
91(100); 92(78)
112(100); 114(33)
91(100); 106(33)
26(49) ; 27(100);
55(64); 56(83)
26(100); 51(32);
52(75); 53(99)
Ion used to
quantify
97
78
106
77
173
164
168
55
92
112
106
56
53
-------
- 16 -
Organics by Liquid-Liquid Extraction -
Gas Chromatography
1. Scope
This method is designed to determine those "unambiguous
priority pollutants" associated with the Consent Decree, that
are solvent extractable and amenable to gas Chromatography.
These compounds are listed in Tables III to V of this section.
Except for the pesticides, it is a gas chromatographic-mass
spectrometric method intended for qualitative and semi-
quantitative determination of these compounds during the
survey phase fof the industrial effluent study. Pesticides
are initially determined by electron capture-gas Chromatography
and, qualitatively, confirmed by mass spectrometry.
2. Special Apparatus and Materials
Separatory funnels - 2 and 4-liter with Teflon stopcock
Continuous liquid-liquid extractors - any such apparatus
designed for use with solvents heavier than water
and having a capacity of 2 to 5-liters . Con-
necting joints and stopcocks must be of Teflon or
glass with no lubrication.
3. Procedure
Sample Preparation for GC-MS Survey - Blend the com-
posite sample to provide a homogeneous mixture including
(a) Available from Aldrich Chemical Co., Milwaukee, WI,
Catalog No. Z10, 157-5.
-------
- 17 -
a representative portion of the suspended solids that are
present. No specific method is required but a motor driven
mechanical stirrer with a propeller type blade is suggested.
Stirring with metal devices is acceptable for organic sampling.
Transfer the sample from the composite container through
j
a glass funnel into a 2-liter graduated cylinder and measure
the volume. Then transfer to a 4-liter separatory funnel or
a continuous extractor as described below. Rinse the cylinder
with several portions of the first volume of extracting sol-
vent. Note: [Either separatory funnel or continuous ex-
traction is acceptable for isolation of the organics. Contin-
uous extraction must be used when emulsions cannot be broken.
See discussion under Emulsions.]
Base-Neutral Extraction
Separatory Funnel Extraction - Adjust the pH of the sample
with 6 N NaOH to 11 or greater. Use multirange pH paper for
the measurement. Serially extract with 250 x 100 x 100 ml
portions of distilled-in-glass methylene chloride. (About 40 ml
of the first 250 ml portion will dissolve in the sample and not
be recovered.) Shake each extract for at least 2 min by the
clock.
Dry and filter the solvent extract by passing it through
a short column of sodium sulfate. Concentrate the solvent by
Kuderna-Danish (K-D) evaporation (distillation). The sodium
sulfate should be prewashed in the column with methylene
-------
- 18 -
chloride. [Note: Check sodium sulfate blank and/ if
necessary, heat in an oven at 500°C for 2 hours to remove
interfering organics.] After drying the extract, rinse the
sodium sulfate with solvent and add to the extract.
Evaporate the extract to 5-10 ml in a 500 ml K-D apparatus
fitted with a 3-ball macro-Snyder column and a 10 ml calibrated
receiver tube. Allow the K-D to cool to room temperature.
Remove the receiver, add fresh boiling chips, attach a two-
chamber micro-Snyder column and carefully evaporate to 1.0 ml
or when active distillation ceases. Remove the micro-Snyder
column and carefully evaporate to 1.0 ml or when active dis-
tillation ceases. Remove the micro-Snyder column and add the
internal standard: 10 yl of 2 vg/yl d,Q-anthracene (per each
ml of extract). Mix thoroughly-
If it is to be overnight or longer before the extract is
run by GC-MS, transfer it from the K-D ampul with a disposable
pipet to a solvent tight container. The recommended container
is a standard 2 ml serum vial with a crimp cap lined with
Teflon coated rubber. These are inert and methylene chloride <
can be held without evaporation loss for months if caps are
unpierced. When the extracts are not being used for analysis,
store them with unpierced caps in the dark and at refrigerator
or freezer temperatures.
Acid (Phenols) Extraction - Adjust the pH of the base-
neutral extracted water with 6 N HC1 to 2 or less. Serially
-------
- 19 -
extract with 200 x 100 x 100 ml portions of distiiled-in-
glass methylene chloride. (Note that only 200 ml is used
for the first extraction). Proceed as described for the base-
neutral extract, including the addition of the internal
standard.
Emulsions - The recovery of 85% of the added solvent
will constitute a working definition of a broken emulsion.
(You may correct the recovery of the first portion for water
solubility of methylene chloride.) Any technique that meets
this criteria is acceptable. Among techniques that have been
tried on these samples with fair success are:
1. Centrifugation of the emulsion layer after removel
of any separated solvent.
2. Passage of the emulsion through a column plugged
with a ball of methylene chloride-wet glass wool.
The solvent used to wet the wool and to wash it
after the emulsion goes through must be measured
and subtracted from the total volume to determine
85% recovery.
3. Relative to labor, solvent is cheap. The addition
of excess solvent sometimes breaks weak emulsions.
You must remember to use excess solvent in the
blanks also,
4. Let the emulsion stand for up to 24 hrs.
5. Draw off the small amount of free solvent that sep-
arates and slowly drip it back in the top of i;he
-------
- 20 -
separatory funnel and through the sample and
emulsion.
Other ideas include stirring with a glass rod, heating
on a steam bath, addition of concentrated sodium sulfate
solution, and sonication. See discussion in Appendix I.
Continuous Extraction - If you cannot achieve 85% solvent
recovery, start with a fresh aliquot of sample and extract by
continuous extraction.
Adjust the pH of the sample as appropriate, pour into
<.
the extractor, and extract for 24 hours. When extracting a
2-liter sample, using the suggested equipment, two liters of
blank water must be added to provide proper solvent recycle.
For operation, place 200-300 ml of solvent in the ex-
tractor before the sample is added and charge the distilling
flask with 500 ml of solvent. At the end of the extraction
remove the solvent from the distilling flask only and evap-
orate and treat as described in the base-neutral extract
section.
Blank Extraction: It is not entirely certain that
2 liters of blank will always be available. When it is,
proceed to process it as the corresponding sample was done.
Include any emulsion breaking steps that used glass wool,
excess solvent or additional chemicals. If less than 2 liters
is available, measure the blank and bring it to volume with
distilled water. On analysis make the necessary quantita-
tive corrections.
-------
- 21 -
Pesticides: These compounds are to be analyzed by
EC-GC using the EPA method published in the Federal Register,
Vol. 38, Number 125, Part II, pp. 17318-17323. (Friday,
June 29, 1973). One-liter rather than 100 ml is to be ex-
tracted. The solvent amounts given in the method and other
parameters remain unchanged. If pesticides are found by EC,
the extract is to be carefully evaporated (clean airstream)
to 0.5 ml and sent for GC-MS confirmation.
The compounds to be analyzed by EC-GC are listed in
Table III.
If the pesticide sample has been received in a 1-gal.
bottle, hand shake the bottle for 1 rain, by the clock to evenly
suspend sediment. Pour the sample into a 1-liter graduated
cylinder and measure the volume. Then transfer the sample
to a 2-liter separatory funnel and rinse the cylinder with
the first volume of extracting solvent. Use additional small
volumes of solvent if necessary to transfer all of the sample.
Proceed with the extraction using the solvents and amounts
prescribed in the published method.
If the sample is to be taken from the original composite
bottle, homogeneously mix as described earlier and transfer
a 1-liter aliquot to a graduated cylinder, then transfer
to the separatory funnel with the aid of a glass funnel 'and
rinse the cylinder as above.
-------
- 22 -
if intractable emulsions are encountered that cannot
be broken as described in the GC-MS survey section/ then a
fresh 1-liter sample should be processed in a continuous
extractor using methylene chloride as the solvent as des-
cribed earlier. The methylene chloride will have to be
evaporated to a small volume and exchanged into hexane for
clean-up or EC-GC analysis. To do this, evaporate the methy-
lene chloride to 6 to 8 ml, cool, add 20 ml of hexane and
a fresh boiling stone and re-evaporate to the desired analy-
tical volume ((5 ml or less) .
Final storage and transport of sample extracts: After
analysis, the extracts of the base-neutrals, acids, blanks
and pesticides are to be sent to ERL, Athens, GA 30601,
ATTN: Dr. Walter Shackelford.
Each extract is to be washed out of its container into
a 10 ml glass ampul and brought to 5 ml ± 1 ml. Methylene
chloride is the solvent for the base-neutrals and acids,
•
hexane for pesticides. The ampuls are to be sealed in a
rounded-off, fire polished manner, i.e., no thin sharp peaks
of glass that are easily broken on handling and shipping.
After sealing the ampuls, put an indelible mark at the
solvent level. Securely attach a label or tag that gives:
Type of fraction (base-neutral, etc.)
Industrial category
Name (of plant, city and state)
-------
- 23 -
Specific source or stage of treatment
Date sampled originally
Date sealed .
i
Name of contractor and analytical laboratory
Wrap the ampuls in packing material to prevent breakage
and mail or ship them postpaid at ambient temperature. When
the samples are safely in ampuls/ the remainder of the com-
posite sample may be discarded.
GC-MS Analysis
Compounds to be analyzed by GC-MS alone fall into two
categories—those in the base-neutral extract (Table iv) and
those in the acid extract (Table V ). Pesticides (Table III
that were tentatively identified in the pesticide analysis
will be confirmed by GC-MS.
The base-neutral extractables may be separated and eluted
into the MS under the following chromatographic conditions:
Column - 6 foot, 2.0 mm inside diameter,- glass
Packing - 1% SP2250 on 100/120 mesh Supel.copbrt
Program - hold 4 minutes @ 50°, program 50°-260°
§ 8°/min., hold 20 minutes @260°
Injector - 275°
Separator - 275°
Carrier gas - He @ 30 ml/min
Injection size - >2 ul
-------
- 24 -
Table IV lists the 49 base-neutral extractable compounds
in order of relative retention times (compared to hexachloro-
benzene) for the above GC conditions. Detection limits were
determined by MS response. The seven compounds without re-
tention times or limits of detection were not available for
this report. It is not recommended that 2,3,7,8-tetrachloro-
dibenzo-p-dioxin (TCDD) be acquired due to its extreme tox-
icity. Based on their similarity to compounds that were avail-
able all seven are expected to be chromatographable using
these standard conditions. In addition the characteristic
masses recommended for MS identification are listed in Table IV.
The limits of detection given in Tables ill and IV refer
to the quantity necessary to inject to get confirmation by
the MS methods described below.
At the beginning of each GC-MS run of a base-neutral
extract, the operator should demonstrate the ability to chro-
matograph benzidine at the 40 ng level. Only after this is
accomplished should the run be started. If benzidine can be
chromatographed, the other nitrogen-containing compounds of
Table IV can be chromatographed as well.
If desired, capillary or SCOT columns may be used instead
of the packed column of SP-2250. Coatings of OV-17 or SP-2250
may be used. The elution order of OV-17 and SP-2250 are very
similar. Some specific data for OV-17 is given in Table VII.
•
The performance criteria for benzidine must still be met
-------
- 25 -
and in addition, the system must be shown to elute the late
running polynuclear aromatic compounds.
The acid extractables may be chromatographed as follows:
Column, 6 foot, 2.0 mm inside diameter, glass
Packing - Tenax GC, 60/80 mesh
Program - 180° - 300° @ 8°/min
Injector - 290°
Separator - 290°
Carrier Gas - He @ 30 ml/min
Injection size - >2 ul
Table V lists the 11 acid extractables in order of
relative retention times (compared to 2-nitrophenol). Chroma-
tography of nitrophenols is poor. The limits of detection
given refer to the amounts required to get MS confirmation
by the methods described below. See Appendix I.
Before an acid extract is run on the GC-MS the operator
should demonstrate the ability to detect 100 ng of>penta-
chlorophenol.
Mass Spectrometry should be conducted with a system
utilizing a jet separator for the GC effluent since membrane
separators lose sensitivity for light molecules and glass
frit separators inhibit the elution of polynuclear aromatics.
A computer system should be interfaced to the mass spectro-
meter to allow acquisition of continuous mass scans for the
duration of the chromatographic program. The computer system
-------
- 26 -
should also be equipped with mass storage devices for
saving all data from GC-MS runs. There should be computer
software available to allow searching any GC-MS run for
specific ions and plotting the intensity of the ions with
respect to time or scan number. The ability to integrate
the area under any specific ion plot peak is essential for
quantification.
To indicate the presence of a compound by GC-MS, three
conditions must be met. First, the characteristic ions for
the compound ,(Tables m-v) must be found to maximize in the
same spectrum. Second, the time at which the peak occurs
must be within a window of ± 1 minute for the retention time
of this compound. Finally, the ratios of the three peak
heights must agree with the relative intensities given in
Tables III-V within ± 20%.
An example of identifying a component is as follows:
It is known that hexachlorobenzene elutes from the SP2250
column at 19.4 minutes. Hexachlorobenzene has characteristic
mass ions at 284(100%), 142(30%), and 249(24%). The computer
is asked to display a plot of the intensities of these ions
versus time (or MS scan number) and the window from 18.4-20.4
minutes is examined for the simultaneous peaking of the in-
tensities of these ions. If all three ions are present, the
ratio of the peak heights is checked to verify that it is
100:30:24 ± 20%. If the three tests are successful, hexachloro-
benzene has been identified in the sample.
-------
- 27 -
Table III lists the 18 pesticides and PCB's that will be
confirmed by GC-MS using the SP2250 column. Chlordane,
toxaphene and the PCB's have retention ranges rather than
specific times due to their being multicomponent mixtures.
It is suggested that the first 14 materials be confirmed
exactly as the other base-neutral compounds.
The last four materials require special treatments. Chicr-
dane is expected to produce two main peaks within the retention
range given in which all three masses listed will maximize.
Toxaphene will produce several (5-15) peaks in which the masses
given will maximize within the retention time range. For the
PCB's each mass given corresponds to the molecular ion of PCB
isomers, e.g., 294 is tetrachlorobiphenyl. A specific mass plot
will show multiple peaks for each of these ions within the re-
tention time listed, but in general they will not maximize in
the same TIC peak. For these four materials in particular it
is necessarv to also run a standard. Because GC-MS is only
being used for confirmation—and at its limit of detection—all
quantification will be done by EC-GC for the pesticides. The
methods for these four are not final and feedback from the
field to Dr. Shackelford is welcome.
When a compound has been identified/ the quantification
of that compound will be based on the integrated area from
the specific ion plot of the first listed characteristic ion
in Tables IV and V. Quantification will be done by the
internal standard method using deuterated anthracene. Response
-------
- 28 -
factors, therefore, must be calculated to compare the MS
response for known quantities of each priority pollutant with
that of the internal standard. The response ratio (R) may
be calculated as:
Ac Ca
R " Aa x Cc
where Ac is the integrated area of the characteristic ion from
the specific ion plot for a known concentration, Cc. Aa and
Ca are the corresponding values for deuterated anthracene.
The relative response ratio for the priority pollutants
should be known for at least two concentration values—40 ng
to approximate 10 ppb and 400 ng to approximate the 100 ppb
level. Those compounds that do not respond at either of these
levels may be run at concentrations appropriate to their res-
ponse. For guidance in MS limits of detection refer to the
values given in Tables III-V.
The concentration of a compound in the extract may now
be calculated using:
_ _ Ac x Ca
C Aa x R
where C is the concentration of a component, Ac is the inte-
grated area of the characteristic ion from the specific ion
plot, R is the response ratio for this component, Aa is the
integrated area of the characteristic ion in the specific
ion plot for deuterated anthracene, and Ca is the concentration
of deuterated anthracene in the injected extract.
-------
- 29 -
In samples that contain an inordinate number of inter-
ferences the chemical ionization (CI) mass spectrum may make
identification easier. In Tables iv and v characteristic
CI ions for most compounds are given. The use of chemical
ionization MS to support El is encouraged but not required.
5. Quality Assurance
GC-MS system performance evaluation is required each day
the system is used for samples or reagent blanks. A sample
of 20 ng of decafluorbtriphenylphosphine is injected into
the system and the mass spectrum is acquired and plotted.
Criteria established in Reference 2 must be met. The analyst
must also demonstrate that the analytical conditions employed
result in sharp total ion current peaks for 40 ng of benzidine
on the SP2250 column when this column is used and 100 ng of
pentachlorophenol on the Tenax GC column when it is used with
the MS as a detector. All plots from the performance evalu-
ation must be retained as proof of valid performance.
As performance evaluation samples become available from
EMSL-Cincinnati, they are to be analyzed by solvent extraction
once each 20 working days and the results reported with other
analytical data.
The 1% SP2250 and Tenax GC column packings are available
by request to EPA contractors from Dr. Walter Shackelford, EPA,
Athens, GA.
(b) Available from PCR, Gainesville, FL
-------
- 30 -
Standards for the priority pollutants may be obtained
from the sources listed in Appendix 1^. Those compounds
marked with an asterisk have not yet been received by the
Athens laboratory.
In order to minimize unnecessary GC-MS analysis of blanks,
the extract may be run on a FID-GC equipped with appropriate
SP2250 and Tenax GC columns. If no peaks are seen of intensi-
ties equal to or greater than the deuterated anthracene internal
standard, then it is not necessary to do a GC-MS analysis. If
such peaks are seen, then the blank must be sent for full
priority pollutant analysis.
The contractor will look for all priority pollutants to
the limit of 10 yg/1 except in those cases listed in Tables iv-V
in which limits of detection are too high for analysis at this
level.
;
6. Reporting of Data
All concentrations should be reported in ranges—10 ppb,
100 ppb, and greater than 100 ppb. Report concentrations for
pesticides as prescribed in the Federal Register Method. The
relative response ratios from MS analysis should be included •
when reporting data. )
All GC-MS data is to be saved on 9-track magnetic tape
and sent to the Athens Environmental Research Laboratory for
storage and later evaluation. The tape format is:
*Those labs which are under contract to perform GC-MS analyzes
for EPA, may obtain a set of standards from Mr. William Telliard,
Chief, Energy'and Mining Branch, (202) 426-2726.
-------
- 31 -
Type - 9 track, 300 BPI, 2400 foot reels
Record length - 80
Block Size - <4QQQ (specify)
Code - EBCDIC
An acceptable data format would have the first two records
containing the sample identification. Subsequent records con-
tain eight mass-intensity pairs, each of which is 10 characters
long. Each mass and each intensity is 5 characters long and
left justified. At the end of each spectrum in a sample run,
the last mass-intensity pair is blank to denote the end of the
spectrum. When all data for the run is on the tape, an end-
of-file mark should be written. The next sample run can then
be entered. One example is:
2 Records:Sample 1 identification
N Records:Spectrum 1 of sample, last mass-intensity
pair is blank to denote end of spectrum
M Records:Spectrum 2 of sample, last mass-intensity
pair is blank to denote end of spectrum
L Records:Spectrum N of sample, last mass-intensity
pair is blank to denote end of spectrum,
END OF FILE
2 Records:Sample 2 identification
etc.
-------
- 32 -
Other data formats are possible, but any format that is
used must be accompanied by a full explanation of all record
formats.
All magnetic tapes, documentation and a table of MS res-
ponse ratios should be sent to:
Dr. W. M. Shackelford
Athens Environmental Research Laboratory
College Station Road
Athens, GA 30601
-------
Table III.
- 33 -
Pesticides
Compound Name
6-endosulfan
a-BHC
Y-BHC
S-BHC•
aldrin
heptachlor
heptachlor epoxide
a-endosulfan
dieldrin
4,4'-DDE
4,4'-DDD
4,4'-DDT
endrin
endosulfan sulfate
6-BHC
chlordane
toxaphane
PCB-1242
PCB-1254
RRT-1-
(hexachlorobenzene)
0
1
1
1
1
1
1
. 1
1
1
1
1
1
1
1
1
0
1
.51
.02
.09
.12
.14
.15
.23
.24
.28
.30
.33
.38
.41
.41
.14-1.37
.22-1.47
.93-1.24
.18-1.41
Detection Limit
(ng)
40
40
40
40
40
40
4u
40
40
40
40
40
40
20
Characteristic
El ions (Rel. Int.)
201(100)
183(100)
183(100)
181(100)
66(100),
100(100)
355(100)
201(100)
79(100),
246(100)
235(100)
235(100)
81(100),
272(100)
, 283(48)
, 109(86)
, 109(86)
, 183(93)
220(11),
, 272(60)
, 353(79)
, 283(48)
263(28) ,
, 248(64)
, 237(76)
, 237(72)
82(61),
, 387(75)
, 278(30)
, 181(91)
, 181(91)
, 109(62)
263(73)
, 274(46)
, 351(60)
, 278(30)
279(22)
, 176(65)
, 165(93)
, 165(59)
263(70)
, 422(25)
183(100), 109(86), 181(90)
373(19), 375(17), 377(10)
(231, 233, 235)*
(224, 260, 294)*
(294, 330, 362)*
* These ions are listed without relative intensities since the mixtures they represent
defy characterization by three masses.
** These three ions are characteristic for the a and y forms of chlordane. No stock
should be set in these three for other isomers.
1% SP-2250 on 100/120 mesh Supelcoport in a 6' x 2 mm" id glass column; He @ 30 ml/min;
Program: 50 for 4 min, then 8 /min to 260 and hold for 15 min.
-------
34 -
Table.IV. Base-neutral Extractables
Compound Name
(hexachloro-
benzene)
1, 3-dichlorobenzene
1,4-dichlorobenzene
foexachloroethane
1,2-dichlorobenzene
bis (2-chloroisopropyl)
ether
hexachlorobutadiene
1,2,4-trichlorobenzene
naphthalene
bis(2-chloroethyl)ether
hexachlorocyclopentadiene
nitrobenzene
bis(2-chloroethoxy)methane
2-chloron aphthalene
acenaphthylene
acenaphthene
isophorone
fJuorene
2,6-dinitrotoluene
1,2-diphenylhydrazine
2,4-dinitrotoluene
N-ni trosod iphenylamine
hexachlorobenzene
4-bromophenyl phenyl ether
phenanthrene
anthracene
dimethylphthalate
diethylphthalate
fluoranthene
pyrene It
di-n-butylphthalate
benzidine
butyl benzylphthalate
0.35
0.36
0.38
0.39
0.47
0.55
0.55
0.57
0.61
0.64
0.64
0.68
0.76
0.83
0.86
0.87
0.91
-0.93
0.96
0.98
0.99
1,
1,
1,
1,
1,
00
01
09
09
10
1.15
1.23
1.30
1-.31
1.38
1.46
Limit of
Detection
(ng)
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40*
40
40*
40
40
40
40
40
40
40
40
40
40*
.40
Characteristic
El ions (Rel. Int.)
146(100),
146(100),
117(100),
146(100),
148(64),
148(64),
199(61),
148(64),
113(12)
113(11)
201(99)
113(11)
45(100)',
225(100,
74(100),
128(100)
93(100),
237(100)
77(100),
93(100),
162(100)
152(100)
154(100)
82(100),
166(100)
165(100)
77(100),
165(100)
169(100)
284(100)
248(100)
178(100)
178(100)
163(100)
149(100)
202(100)
202(100)
149(100)
184(100).
149(100)
77CU9), 79(12)
223(63), 227(65)
109(80), 145(52)
, 127(10), 129(11)
63(99), 95(31)
, 235(63), 272(12)
123(50), 65(15)
95(32), 123(21)
, 164(32), 127(31)
, 153(16), 151(17)
, 153(95), 152(53)
95(14), 138(18)
, 165(80), 167(14)
, 63(72), 121(23)
93(58), 105(28)
, 63(72), 121(23)
168(71)
142(30),
250(99),
179(16),
179(16),
164(10),
178(25),
101(23),
101(26),
150(27),
167(50)
249(24)
141(45)
176(15)
176(15)
194(11)
150(10)
100(14)
100(17)
104(10)
,92(24),
, 91)50)
185(13)
CI ions
(Methane)
146, 148, 150
146, 148, 150
199, 201, 203
146, 148, 150
77, 135, 137
223, 225, 227
181, 183, 209
129, 157, 169
63, 107, 109
235, 237, 239
124, 152, 164
65, 107, 137
163, 191, 203
152, 153, 181
154, 155, 183
139, 167, 178
166, 167, 195
183, 211, 223
185, 213, 225
183, 211, 223
169, 170, 198
284, 286, 283
249, 251, 277
178, 179, 207
178, 179, 207
151, 163, 164
177, 223, 251
203, 231, 243
203, 231, 243
149, 205, 279
185, 213, 225
149, 299, 327
-------
- 35 -
Table -'IV. Base-neutral Extractables (Cont'd.)
Compound Name
(hexachloro-
benzene)
chrysene
bis (2-ethylhexyl)phthalate
benzo(a)anthracene
benzo (b)fluoranthene
benzo(k)fluoranthene
benzo(a)pyrene
indeno(1,2,3-cd)pyrene
dibenzo(a,h)anthracene
benzo(g h i)perylene
46
50
54
66
66
73
07
2.12
2.18
N-nitrosodimethylamine
N-nitrosodi-n-propylamine
4-chloro-phenyl phenyl ether
endrin aldehyde
3,3' -dichlorobenzidine
2,3,7,8-tetrachlorodibenzo-
p-dioxin
bis (chioromethyl)ether
deuterated anthracene (dlO) 1.09
Limit of
Detection
(ng)
40
40
40
40
40
40
100
100
100
Characteristic
El ions (Rel. Int.
40
228 (100)
149(100)
228,(100)
252(100)
252(100)
252(100)
276 (100)
278(100)
276(100)
42 (100) ,
130(22) ,
204(100)
252(100)
322(100)
45(100),
188(100)
, 229(19)
, 167(31)
, 229(19)
, 253(23)
, 253(23)
, 253(23)
, 138(28)
, 139(24)
, 138(37)
74(88), t
42(64), ]
, 206(34)
, 254(66)
, 320(90)
49(14) ,
, 94(19),
226(23)
279(26)
226 (19)
125(15)
125(16)
125(21)
277(27)
279(24)
277(25)
14(21)
L01(12)
, 141(29)
, 126(16)
, 59(95)
51(5)
80(18)
CI ions
(Methane)
228, 229, 257
149 '
228, 229, 257
252, 253, 281
252, 253, 281
252, 253, 281
276, 277( 305
278, 279, 307
276, 277, 305
189, 217
1% SP-2250 on 100/120 mesh Supelcoport in a 61 x 2 mm id glass column; He @ 30 ml/min;
Program: 50° for 4 min, then 8 /min to 260 and hold for 15 min.
Conditioning of column with base is required. - -
-------
- 36 -
Table V. Acid Extractables
Compound Name
RRT
(2-nitrophenol)
2-chlorophenol , . ,0..63
phenol 0.66
2,4-dichlorophenol 0.96
2-nitrophenol 1.00
p-chloro-m-cresol 1.05
2,4,6-trichlorophenol ' 1.14
2,4-dimethylphenol 1.32
2,4-dinitrophenol 1.34
4,6-dinitro-o-cresol 1.42
4-nitrophenol 1.43
pentachlorophenol 1.64
deuterated anthracene (dlO) 1.68
Limit of
Detection
(ng)
100
100
100
100
100
100
100
2 yg
2 vi g
100
100
40
Characteristic
El ions (Rel. Int.)
128(100), 64(54), 130(31).
94(100), 65(17), 66(19)
162(100), 164(58), 98(61)
139(100), 65(35) , 109(8)
142(100) , 107(80), 144(32)
196(100), 198(92), 200(26)
122(100), 107(90), 121(55)
184(100), 63(59) , 154(53)
198(100), 182(35), 77(28)
65(100), 139(45) , 109(72)
266(100), 264(62), 268(63)
188(100), 94(19), 80(18)
CI ions
(Methane)
129, 131,^157
95, 123, 135
163, 165, 167
140, 168, 122
143, 171, 183
197, 199, 201
123, 151, 163
185, 213, 225
199, 227, 239
140, 168, 122
267, 265, 269
189, 217
Column: 6' glass, 2 nun i.d.
Tenax GC - 60/80 mesh
180° - 300° 6 8°/min.
He <§ 30 ml/ruin
-------
- 37 -
Table VI. ELUTION ORDER OF MOST OF THE SEMIVOLATILE
PRIORITY POLLUTANTS ON 1% SP2250a
Compound RRT 'C
1,3-dichlorobenzene 0.35
2-chlorophenol 0.35e
1,4-dichlorobenzene 0.36
hexachloroethane 0.38
1,2-dichlorobenzene 0.39
bis(2-chloroisopropyl)ether • 0.47
3-endosulfan 0.51
2,4-dimethyl, phenol 0.52e
2-nitrophenol 0.53e
2,4-dichlorophenol 0.53e
hexachlorobutadiene 0.55
1,2,4-trichlorobenzene 0.55
naphthalene 0.57
bis(2-chloroethyl)ether 0.61
hexachlorocyclopentadiene 0.64
nitrqbenzene 0.64
phenol 0.67
bis(2-chloroethoxy)methane 0.68
2,4,6-trichlorophenol 0.71e
p-chloro-m-cresol 0.73
2-chloronaphthalene 0.76
acenaphthylene 0.83
acenaphthene 0.86
isophorone 0.87
fluorene 0.91
-------
- 38 -
Table VI. ELUTIOM ORDER OF MOST OF THE SEMtVOLATlLE
PRIORITY POLLUTANTS ON 1% SP22503 (Continued)
b c
Compound ^ RRT ' _
2,6-dinitrotoluene 0.93
1,2-diphenylhydrazine 0.96
2,4-dinitrotoluene 0.98
N-nitrosodiphen,ylamine 0.99
hexachlorobenzene 1.00
4-brorcophenyl phenyl ether 1.01
a-BHC 1.02
f
Y-BHC " 1.09
phenanthrene 1.09
anthracene , 1.09
dimethyl phthalate ' 1.10
f
pentachlorophenol 1.11
S-BHC 1.12
aldrin 1.14
diethyl phthalate 1.15
heptachlor 1.15
heptachlor epoxide 1.23
fluoranthene 1*23
ct-endosulfan 1.24f
dieldrin 1428;
4,4'-DDE • 1.30
pyrena ' 1*30
di-n~butyl phthalate 1.31
4,4'-DDD (p,p'-TDE) 1.33
4,4'—DDT 1.38d
endosulfan sulfate 1.41
endrin . 1.41
benzidine - 1.38
butyl benzyl phthalate 1.46
chrysene 1.46
-------
- 39 -
Table VI. ELUTION ORDER OF MOST OF THE SEMI VOLATILE
PRIORITY POLLUTANTS ON 1% SP22503 (Continued)
Coirtpound
•KRTbj
bis(2-ethylhexyl)phthalate 1.50
benzo(a)anthracene 1.54
benzo(b)fluoranthene 1.66
benzo(k)fluoranthene 1.66
benzo(a)pyrene 1.73
indeno(l,2,3-cd)pyrene 2.07
dibenzo(a,h)anthracene 2.12
benz o(ghi)pe rylene 2.12
a 1% SP-2250 on 100/120 mesh Supelcoport in a 6f x 2irtm id
glass column; He @ SOml/min; Program: 50° for 4 min,
then 8°/min to 260° and hold for 15 min.
Relative to hexachlorobenzene at 19.4 min.
40ng gives 5-90% response on FID unless otherwise noted.
200ng required to obtain 5-90% response on FID.
2 11 g required.
40 yg required.
-------
- 40 -
Table 'VI
(continued)
Standards not available; as of 2/8/77
N-nitrosodi-n-propylamine
4-chlorophenyl phenyl ether
TCDD
endrin aldehyde
N-nitrosodimethylaraine
3,3' -dichlorobenzidine
bis (chlorome thy 10 ether (unstable in water)
Standards that would not chromatograph;
4 ,6-dinitro-o-cresol
4 -ni tropheno 1
2 , 4-dinitrophenol
Standards yielding a range of eaks;
PCB-1242 0.93-1.24
PCB-1254 1.18-1.41
toxaphene- 1 . 22-1 . 4 7
chlordane 1.14-1.37
-------
- 41 -
- .._ Table VII. Order of Elution for
OV-17 SCOT Column
2
Compound Spectrum Number
1,3-dichiorobenzene 134
1,4-dichlorobenzene 137
2-chlorophenol 141
1,2-dichlorobenzene ' 153
bis(2-chloroethy1)ether 163
phenol 165
bis (2-chloroisopropyl)ether 173
hexachloroethane 178
nitrobenzene ' 194
2-nitrophenol 219
1,2,4-trichlorobenzene 234
2,4-dimethylphenol 240
naphthalene 240
2,4-dichlorophenol 244
hexachlorobutadiene 262
isophorone 272
p-chloro-m-cresol 317
hexachlorocyclopentadiene 325
2,4,6-trichlorophenol 332
chloronaphthalene 339
2,4-dinitrotoluene 372
acenaplithylene 374
acenaphthene 390
dimethylphthalate 397
fluorene 434
diethylphthalate 447
N-nitrosodiphenylamine 447
2,6-dinitrotoluene ^ 454
a-BHC ' 476
4-bromophenyl phenyl ether 478
T-BHC 487
hexachlorobenzene 490
8-BHC . 506
phenanthrene 518
anthracene 518
di-n-butylphthalate 583
aldrin 592
fluoranthene ' 617
pyrene 634
DDE 659
ODD • 664
endrin 688
dieldrin 688
DDT . 713
butyl benzyl phthalate 713
benzo(a)anthracene 748
chrysene 748
-------
- 42 -
Table VII. • continued
2
Compound Spectrum Number
bis(2-ethylhexyl)phthalate 804
benzo(a)pyrene 906
banzo(b)fluoranthene 970
benzo(k)fluoranthene 970
33 meter glass OV-17 SCOT column,
Program: 60 -260 @6 /minute
2
Number of 2.5 second scans up to point of elution.
-------
- 43 -
Metals
1. Sample Preparation
With the exception of mercury, the metals to be deter-
mined may be divided into two groups as follow:
a) those metals which are to be first analyzed by
flame atomic absorption (AA), and, if not detected,
then analyzed by flameless AA—Be, Cd, Cr, Cu, Ni,
Pb and Zn,
b) those metals which are to be analyzed by flameless
AA only—Ag, As, Sb, Se, and Tl.
For flame AA analysis the sample should be prepared using
the procedure as given in "Methods for Chemical Analyses of
Water and Wastes (1974)", 4.1.4, page 83 (Reference 7).
With the exception of antimony and beryllium, samples to
be analyzed by flameless AA should be prepared as an industrial
effluent as described in "Atomic Absorption Newsletter," 14,
page 111 (1975) (Reference 8). Note: Nickel nitrate should
be added only to those aliquots on which the analysis of
selenium and arsenic are to be accomplished. The sample prep-
aration procedure for antimony and beryllium analysis by flame-
less AA is the same procedure used for flame AA.
The sample preparation procedure to be used for mercury
analysis is that given in "Methods for Chemical Analysis of
Water and Wastes (1974)", 8.1, page 124 (Reference 7).
-------
- 44 -
2. Apparatus
All samples are to be analyzed using an atomic absorption
spectrophotometer equipped with simultaneous background
capability. For arsenic, cadmium, antimony, selenium, thallium,
and zinc, either electrodeless discharge lamps or high intensity
hollow cathode lamps may be utilized. A heated graphite atom-
izer is to be used for all flameless AA work. A strip chart
recorder must be used as part of the readout system to detect
and avoid the inclusion of extraneous data.
3. Procedure
a) Flame AA - The procedures to be used are those
described in "Methods for Chemical Analysis of
Water and Wastes (1974)"(Reference 7) as referenced
in Table I below. Instructions as to when flame-
less AA is to be used are also included. For
those defined in the recommended procedures, the
instrument manufacturers recommendations are to
i
be followed. Background correction is to be used
on all analyses.
-------
- 45 -
Table VIII
Methods for Chemical
Analysis of Water and
Element Wastes, 1974* Comments
Be p. 99 Analyze by flameless AA if
cone. <20 yg/1
Cd p. 101 Analyze by flameless AA if
cone. <20 yg/1
Cr p. 105 Use nitrous oxide-acetylene
flame for all analyses—analyze
by flameless AA if cone. <200 yg/1
Cu p. 108 Analyze by flameless AA if
cone. <50 yg/1
Ni p. 141 Analyze by flameless AA if
cone. <100 yg/1
Pb p. 112 Analyze by flameless AA if
cone. <300 yg/1
Zn p. 155 Analyze by flameless AA if
cone. <20 yg/1
*In those instances where more vigorous digestion for sample
preparation is desired (or necessary) the procedure on page 82
(4.1.3) should be followed.
b) Standard solutions to be used for the flameless
work should also be prepared as described in
"Methods for Chemical Analysis of Water and Wastes
(1974)" (Reference 7). The working standards should
be diluted to contain the same acid concentration as
the prepared samples. The instrumental settings
and conditions recommended by the manufacturers are
to be considered the procedural guidelines. In
addition, the following requirements should also be
incorporated into the procedures:
-------
- 46 -
1) Argon should be used as the purge gas in
,
all analyses.
2) Background correction and method of standard
addition must be used on all analyses.
3} A blank maximum temperature atomization, without
gas interrupt, should be accomplished before
each analytical determination.
4) The graphite tube or cuvette should be replaced
as suggested by the instrument manufacturer or
when contamination or lack of precision indicates
that replacement is necessary.
5) All disposable pipet tips should be cleaned
before use by soaking overnight in 5% redistilled
nitric acid, rinsed with tap and deionized
water, and dried.
6) The accuracy of the temperature indicator on the
heated graphite atomizer should be verified
before beginning any analytical work. This
should be done by plotting charring temperature
for a standard solution of a compound where the
volatilization temperature is known. The com-
pound used should have a volatilization temper-
ature between 800 and 1200°C.
7) To insure that there is no loss from the acid ,
i
matrix prior to atomization, the optimum charring
temperature for each metal should be established
in the same manner (i.e., by plotting charring
temperature versus atomization signal of standard
solution of each metal).
-------
- 47 -
For the determination of selenium the procedure given
for industrial effluents ("Atomic Absorption Newsletter,"
Vol. 14, page 109 [1975]) (Reference 8) should be followed.
Arsenic should be determined in the same manner (using the
nickel nitrate matrix) with an optimum charring temperature
of approximately 1300°C.
The analysis of zinc by flameless AA is difficult because
of environmental contamination. The analyst must take pre-
caution to provide a clean work area to minimize this problem.
c) Mercury analyses - The cold vapor technique as
described in "Methods for Chemical Analysis of Water
and Wastes, (1974)", page 118 (Reference 7) is to
be followed.
4-. Quality Assurance
a) To verify that the instrument is operating correctly
within the expected performance limits, an appropriate
standard should be included between every ten samples.
b) Spiked aliguots shall be analyzed with a frequency
of 15% of the sample load for each metal determined
by flame AA. If the recovery is not within ±10% of
the expected value the sample should be analyzed by
method of standard addition. (The spike should be
added to the aliquot prior to sample preparation.)
The amount added should increase the absorbance by
not less than 0.01 units where the absorbance in the
unspiked aliquot was less than 0.1, and not more than
0.1 when the absorbance in the unspiked aliquot was
0.1 or greater.
-------
- 48 -
c) For mercury, the spike added should be an
amount equal to five times the detection level.
d) Reagent blanks shall be run for each metal
being determined with the sample values being
corrected accordingly.
e) When using the method of standard addition, a
linear curve over the entire range of addition
is necessary for the results to be considered
valid.
5. Data Reporting
Report all metal concentrations as follows: Less than
10 ug/1, nearest yg; 10 yg/1 and above, two significant figures
-------
- 49 -
Cyanides
1. Sample Preparation
All samples are to be distilled prior to determination for
total cyanides. The distillation procedure given on page 43
of "Methods for Chemical Analysis of Water and Wastes, (1974)"
(Reference 7) is to be followed.
2. Sample Procedure
The procedure for total cyanides as given on pages 43-48
of "Methods for Chemical Analysis of Water and Wastes, (1974)"
(Reference 7) is to be followed.
3. Quality Assurance
a) Initially, demonstrate quantitative recovery with
each distillation-digestion apparatus by comparing
distilled standards to non-distilled standards.
Each day, distill at least one standard to confirm
distillation efficiency and purity of reagents.
b) At least 15% of the cyanide analysis will consist
of duplicate and spiked samples. Quality control
limits are to be established and confirmed as described
in Chapter 6 of the "Analytical Quality Control
Handbook" (Reference 9) .
4. Reporting of_ Data
•
Report cyanide concentrations as follows: Less than
1.0 mg/1, nearest 0.01 mg; 1.0 mg/1 and above, two significant
figures.
-------
- 50 -
Phenols
1. Sample Preparation
c
Distill all samples prior to determination of phenols.
Use the procedure in "Standard Methods for the Examination of
Water and Wastewater," 14th edition, 1975, p. 576 (Reference 10),
2. Procedure :
Use method 510 for phenols in Appendix Xf pages 577-580
and 580-581. Use method 510B for samples that contain less
than 1 mg/1 of phenol. Use method 510C for samples that contain
more than 1 mg/1 of phenol.
3. Quality Assurance
Demonstrate quantitative recovery with each distillation
apparatus by comparing distilled standards to non-distilled
standards. Each day distill, at least, one standard to con-
firm the distillation efficiency and purity of reagents.
Run duplicate and dosed sample analyses on at least 15%
of the samples analyzed for phenol. Establish and confirm
quality control limits as described in Reference 9.
4. Reporting of Data
Report phenol concentrations as follows:
Method 510B to the nearest yg/1.
Method 510C - when less than 1.0 yg/1 to the nearest
•
0.01 mg; 1.0 mg/1 and above to two significent figures.
Report all quality control data when reporting results
of sample analysis.
-------
- 51 -
REFERENCES
1. Determining Volatile Organics at Microgram-per-Liter Levels
by Gas Chromatography. T. A. Bellar and J. J. Lichtenberg,
Jour. AWWA, p. 739-744, Dec. 1974.
2. Reference Compound to Calibrate Ion Abundance Measurements
in Gas Chromatography—Mass Spectrometry Systems. J. W.
Eichelberger, L. E. Harris and W. L. Budde, Anal. Chem. 47,
995-1000 (1975).
3. ASTM Annual Standards - Water, part 31, Method D2908 "Standard
Recommended Practice for Measuring Water by Aqueous-Injection
Gas Chromatography."
4. ASTM Annual Standards - Water, part 31, Method D3371 "Tentative
Method of Test for Nitriles in Aqueous Solution by Gas Liquid
Chromatograph."
5. Harris, L. E., Budde, W. L. and
Eichelberger, J. W., Anal. Chem., 46, 1912
(1974). "Direct Analysis of Water Samples for Organic Pollu-
tants with Gas Chromatography-Mass Spectrometry-"
6. Federal Register, Volume 38, number 125, part II, Appendix II,
p. 17319, Friday, June 29, 1975, "Determination of Organo-
chlorine Pesticides in Industrial Effluents,"
7. Methods for Chemical Analysis of Water and Wastes (1974).
U.S. Environmental Protection Agency, Technology Transfer.
8. Determining Selenium in Water, Wastewater, Sediment and Sludge
by Flameless Atomic Absorption Spectroscopy. T. D. Martin and
J. F. Kopp, Atomic Absorption Newsletter 14, 109-116 (1975).
-------
. - 52 -
9. Handbook for Analytical Quality Control in Water and Waste-
water Laboratories (1972). U.S. Environmental Protection
Agency, Technology Transfer.
10. "Standard Methods for the Examination of Water and Waste-
water ," 14th edition, 1975.
-------
- 53 -
APPENDIX I
General Information
Emulsions
Limited work with several categories of industrial effluents
covered by this study (tanneries, petroleum, soap and detergent,
steam electric, pesticide) show that emulsions of widely differing
frustration factors are often encountered in the extraction pro-
cedure. Samples that emulsify at basic pH usually also emulsify
at acid pH. There are two equally acceptable alternatives avail-
able for the purposes of this protocol: break the emulsion or start
over with fresh sample and use a continuous extractor, to prevent
the formation of emulsions.
By the 85% solvent recovery criteria, no way was found to break
the emulsion formed on extraction of untreated tannery wastes. A
soap and detergent sample was also very difficult. The use of a
continuous heavier-than-water liquid extractor allowed the extraction
to take place with no difficulties and very little labor. However,
two days time is required. Comparison of samples from four industries-
petroleum, tannery, pesticide, and soap and detergent—by both shake-
out and continuous extraction using wastes spiked with priority pollu-
tants indicate that the two techniques are comparable. For some
individual cases one technique is better than the other but no clear
pattern emerges. Therefore, if desired, a continuous extraction
technique may be used in place of sepa-ratory funnel extraction for
all samples as well as those for which it is absolutely necessary
because of intractable emulsions.
-------
- 54 -
APPENDIX I
(continued)
There is a justifiable concern that the extraction efficiency
for these compounds may differ widely depending on the nature of
the effluents. This is true but no better approach is apparent.
For example, recoveries of most of the base-neutrals were judged
to be about 75% from the tannery and petroleum samples but less
than 25% from soap and detergent.
Acid (Phenol) Analysis
Although the 11 phenols of interest here do chromatograph on
the Tenax column cited, the chromatography is poor, particularly
for the nitrophenols. Two other columns have shown good response
for the acid extractables. SP2250 can be used for this purpose.
Phenol responses on SP2250 are shown in Table IV. It should be
noted, however, that 4-nitrophenol, 2,4-dinitrophenol, 4,6-dinitro-
o-cresol, and pentachlorophenol failed to give MS response at the
100 ng level using this column.
SP1000 (4% load) has also been evaluated for use with the
acid fraction. All but 2,4-dinitrophenol and 4,6-dinitrc-o-cresol
elute from this column. Pentachlorophenol and 4-nitrophenol are
eluted from SP1000, but they produce broad peaks which are difficult
to quantify.
-------
- 55 -
Appendix II
Possible Sources for Some Priority Pollutant Standards
Source of
Compound Standard 2
acenaphthene AN p. 118
acrolein AL p. 18
acrylonitrile AL p. 19
aldrin HERL #30
dieldrin HERL §2380
benzene B p. 154
benzidine1 " " RFR
carbon tetrachloride (tetrachloromethane) B p. 88
chlordane (technical mixture & metabolites) HERL f!200
Chlorinated benzenes (other than dichlorobensenes)
chlorobenzene AL p. 165
1,2,4-trichlorobenzene AL p. 710
hexachlorobenzene AL p. 416
Chlorinated ethanes (including 1,2-
dichloroethane, 1,1 /1-trichloroethane
and hexachloroethane)
1,2-dichloroethane . AL p. 261
1,1,1-trichloroethane B p. 309
hexachloroethane AL p. 416
1,1-dichloroethane ' PB p. 142
1,1,2-trichloroethane PB p. 388
1,1,2,2-tetrachloroethane 'PB p. 372
chloroethane EA p. 53
Chloroalkyl ethers (chloromethyl, chloroethyl and
mixed ethers)
bis(chloromethyl) ether1 EFRi
bis(2-chloroethyl) ether AL p. 173
2-chloroethyl vinyl ether AL p. 174
Chlorinated naphthalene
•
2-chloronaphthalene ICN p. 50
-------
- 56 -
Appendix II
Possible Sources for Some Priority Pollutant Standards
(Continued)
Compound
Source of
Standard2
Chlorinated phenols (other than those listed
elsewhere;includes trichlorophenols and
chlorinated cresols)
2,4,6-trichlorophenol
p-chloro-m-cresol
chloroform (trichloromethane)
2-chlorophenol
DDT and, metabolites
4,4'-DDT
4,4'-DDE
4,4f-DDD (p,p'-TDE)
V
Dichlorpbenzenes (1,2-;1,3-; and 1,4-
dichlorobenzenes)
1,2-dichlorobenzene
1,3-dichlorobenzene
1,4-dichlorobenzene
Dichlorobenzidine
3,3*-dichlorobenzidine*
Dichloroethylenes (1,1-dichloroethylene and
1,2-dichloroethylene)
1,1-dichloroethylene
1,2-trans-dichloroethylene
2,4-dichlorophenol
Dichloropropane and dichloropropene
1,2-dichloropropane
1,3-dichloropropylene (1,3-dichloropropene)
2,4-dimethylphenol
Dinitrotoluene
2 ,4-dinitrotoluene
2,6-dinitrotoluene
1,2-diphenylhydrazine
AL p. 712
TCI p. 102
B p. 92
AL p. 187
HERL 51920
HERL *1860
HERL 11780
AL p. 258
AL p. 258
AL p. 258
CPL p. 81
AL p. 746
AL p. 262
AL p. 265
AL p. 267
AL p. 267
AL p. 323
PB p. 180
PB p. 180
AL p. 338-
-------
- 57 -
Appendix II
Possible Sources for Some Priority Pollutant Standards
(Continued)
Compound
Source of
Standard 2
Endosulfan and metabolites
ct-endosulfan
g-endosulfan
endosulfan sulfate
Endrin and metabolites
endrin
endrin aldehyde
HERL .£3220
HERL §3200
NI p. 45
HERL §3260
NI p. 147
ethylbenzene
fluoranthene
Haloethers (other than those listed elsewhere)
4-chlorophenyl phenyl ether (p-chloro-
diphenyl ether)
4-bromophenyl phenyl ether
bis(2-chloroisopropyl) ether
bis(2-chloroethoxy) methane
Halomethanes (other than those listed elsewhere)
methylene chloride (dichloromethane)
methyl chloride (chloromethane)
methyl bormide (bromomethane)
bromoform (tribromomethane)
dichlorobromomethane
trichlorofluoromethane
dichlorodifluoromethane
chlorodibromomethane
Heptachlor and metabolites
heptachlor
heptachlor epoxide
hexachlorobutadiene
Hexachlorobyclohexane (all isomers)
cx-BHC
6-BHC
y-BHC (lindane)
6-BHC
B p.
AN p.
161
118
RFR p. 6*
ICN p. 37
PB
PB p. 62
PB
PB
PB
PB
CO
PB
PB
CO
P-
P-
P.
P.
P.
P.
P.
P.
276
277
276
73
16
358
142
27
HERL #3860
HERL #3880
AL p. 416
HERL #620
HERL #640
HERL #680
HERL #660
-------
- 58 -
Appendix n
Possible Sources for Some Priority Pollutant Standards
(Continued)
Compound
Source of
Standard 2
hexachlorocyclopentadiene
isophororie
naphthalene
nitrobenzene
Nitrophenols (including 2,4-dinitrophenol and
dinitrocresol)
2-nitrophenol
4-nitrophenol
2,4-dinitrophenol
4,6-dinitro-o-cresol
Nitrosamines
N-nitrosodimethylamine*
N-nitrosodi-n-propylamine
N-nitrosodiphenylamine
pentachlorophenol
phenol
Phthalate esters j
bis (2-ethylhexyl) phthalate
butyl benzyl phthalate
di-n-butyl phthalate
diethyl phthalate
dimethyl phthalate
Polychlorinated biphenyls (PCB's)
PCB-1242 (Arochlor 1242)
PCB-1254 (Arochlor 1254)
Pqlynuclear aromatic hydrocarbons (including
benzanthracenes, benzopyrenes, benzo-
fluoranthene, chrysenes, dibenzanthracenes,
and indenopyrenes)
1,2-benzanthracene
benzo [a]pyrene (3,4-benzopyrene)
3 r4-benzofluoranthene
11,12-benzofluoranthene
chrysene
AL p. 416
AL p. 464
AN p. 118
AL p. 557
AL p. 564
AL p. 564
AL p. 332
TCI p. 188
NI p. 173
PB p. 310
EA p. 159
AL p. 587
AL p. 595
CS p,
CS p.
CS p.
CS p,
CS p.
8
8
8
8
8
HEKL #5703
HERL 15705
AN p.
AN p.
NI
NI
AN p.
118
118
118
-------
- 59 -
Appendix n
Possible Sources for Some Priority Pollutant Standards
(Continued)
Compound
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
tetrachloroethylene
toluene
toxaphene
trichloroethylene
vinyl chloride (chloroethylene)
1-bromodecane (possible internal standard)
1-bromododecane (possible internal standard)
Source of
Standard 2
acenaphthylene
anthracene
1 , 12-benzoperylene
fluorene
phenanthrene
1,2:5 , 6-dibenzanthracene
indeno (1,2 , 3-C,D) pyrene
pyrene
AN p.
AN p.
AN p.
AN p.
AN p.
AN p.
AN p.
AN p.
1
118
118
118
118
118
118
118
NI p. 174
AL p. 680
AL p. 701
HERL #6740
AL p. 711
PB p. 406
Footnotes :
These compounds or any mixture containing 1% or more by weight
of these compounds are defined as carcinogens in the Federal
Register, Vol. 38, No. 144, pp. 20074-20076, 27 July 1973.
Prescribed safety regulations for handling are in the Federal
Register, Vol. 39, No. 20, pp. 3756-3797, 29 January 1-974.
i
Only one source is listed even though several may be available.
These sources are not to be interpreted as being endorsed by
the EPA; they serve to show at least one ven--'.or where each
standard can be obtained. When several sources were available
and compound purity was listed, the source having the highest
purity material was selected.
These compounds have been ordered but have not been received
at Athens ERL as yet.
-------
- 60 -
Sources of Standards and Abbreviations
AL Aldrich Chemical Co., Milwaukee, Wise.; Catalog 1977-1978.
AN Analabs, Inc., North Haven, Conn.; Catalog 18 (June 1976).
B J. T. Baker Chemical Co., Phillipsburgh, N.J.;
Catalog 750 (July 1975).
CS Chem-Service, West Chester, Pa.; Bulletin CS-100-8 (1975).
CPL Chemical Procurement Laboratories, College Point, N.Y.;
1975 catalog.
EA Eastman Kodak Co., Rochester, N.Y.; Catalog 48 (1976).
ICN K&K Rare & Fine Chemicals, Plainview, N.Y.; Catalog No. 10
(1975).
NI Nanogens International, P.O. Box 487, Freedom, CA 95019
"Nanogen Index" (1975).
PB Pfaltz & Bauer Chemical Co., STamford, Conn.; Catalog
1976.
RFR RFR Corp., Hope, R.I.; "Chemical Standards for Air-Watsr-
Industry-Foods" (1975).
HERL "Analytical Reference Standards and Supplemental Data for
Pesticides and Other Selected Organic Compounds", EPA-
660/9-76-012 (May 1976), Health Effects Research Laboratory,
Environmental Toxicology Division, Research Triangle Park,
NC. A sample order blank for standards and the above •
publication are attached.
t
CO Columbia Organics Catalog A-7, Columbia, S.C. (1975).
TCI Tridom Chemical Inc., Haut-tauge, N.Y. , Catalog No. 1
(1976).
*Those labs which are under contract to perform GC-MS analysis
for EPA, may obtain a set of standards from *fr. William A.
Telliard, Chief, Energy and Mining Branch, Effluent
Guidelines Division, (WH 552) 401 M Street, S.W.
Washington, D.C. 20460 (202) 426-2720.
-------
- 61 -
ENVIRONMENTAL TOXICOLOGY DIVISION
HEALTH EFFECTS RESEARCH LABORATORY
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Research Triangle Park, North Carolina '27711
SUBJECT- Index of Pesticides Analytical Reference
Standards - Update of Hailing List
DATE: June, 1976
FROM:
TO:
Health Effects Research Laboratory, ETD, ACB,
Research Triangle Park, NC, U.S.A. 27711 (MD-69)
l"
All Laboratory Facilities on our Mailing List
This copy of the 1976 revision of our pesticides reference standards
index was mailed to the address appearing on our mailing list. As this
list is several years old, we are sure that a number of addresses have
changed and that some are probably no longer existent.
If you wish to remain on our mailing list to receive future updates
of this publication, would you be good enough to' complete the mail-back
below, snip it off, and return it to us. Do_ not tear off the back cover
to return to the address shown. If you have no use for this publication
but know of some other individual within your organization who is con-
cerned with pesticides analysis, would you convey this index, along with
the mailback, to that person.
To: U. S. Environmental Protection Agency
Health Effects Research Laboratory
Environmental Toxicology Division
Research Triangle Park, NC, U.S.A. 27711
pate
>-69)
D
D
D
We wish to be retained on your mailing list to receive future updates
of the Pesticides Standards Index. The address shown on the envelope
is entirely correct and requires no changes.
We have no interest in future updates of this publication.
cancel us from your mailing list.
Please
We wish to be retained on your mailing list, but the address shown
on the envelope should be changed to read (print or type)
-------
- 62 -
REQUEST FOR ANALYTICAL REFERENCE STANDARDS
TO: Quality Assurance Section
Environmental Toxicology Division,
EPA, HERL, Research Triangle Park, NC 27711
MD-69
Date
Date Request Reed
Shipment Date
Lab Code No.
Order Filled by
DO NOT WRITE IN THIS SPACE
The following reference standards are required for our program:
Catalog Compound H Catalog Compound
Code (Catalog Name) n Code (Catalog Name)
No. | No.
1
h
i
1
1
1
If necessary, use back of sheet to complete list. Covering letter unnecessary
if this form is completed in full.
Name and address of laboratory
Requestor's Name (Print or type)
IMPORTANT:
1. The amount of each standard is restricted to 100 mg because of the scarcity
and expense of refining analytical grade materials.
2. Please return at once, the acknowledgement card enclosed with each shipment.
This provides the sole evidence of delivery of the shipment.
3. Do not request compounds not listed in the catalog. No others are stocked.
4. If a bottle appears to be empty, remove cap and examine interior of bottle
and cap. Certain highly viscous materials tend to collect in cap.
-------
. , APPENDIX III
..."1"~ COLLECTION OF SAMPLES FOR SCREENING ANALYSES
The Initial characterization (screening) of the varied industrial
discharges covered by this program vn"!l be made on an analysis' of
a composite effluent sample. Any scheme for collecting a
composite sample is, in effect a method for mechanically
Integrating to obtain average characteristics of a discharge.
During the screening phase the sample composite can be used to
determine the average characteristics which would be
representative of that discharge. Simple composite samples are
those that are made up of a series of aliquots of constant volume
collected at regular time intervals in a single container. Some
situations may require flew or time proportional sampling, this
determination will be made by the individual project officer
after considering his specific industrial category.
The determination of compositing period 24, 48 or 72 hours will
be made on a case by case basis. The duration of compositing
will depend on the type of sample being collected, the type of
facility being sampled and the time varying characteristics of
the discharge. The rate of change of flow and other
characteristics of the discharge and the accuracy required will
also influence tha determination of the compositing period. For
example longer compositing periods would be warranted when less
•stable unit process operations are being sampled.
Collection of Samples
1. Collection of Composite Samples for Liquid-Liquid Extraction
Collect a representative composite sample. The maximum time
interval between aliquot samples shall be no longer than 30
minutes. The minimum aliquot size shall be ICO ml. •/ The
sample must be collected with an automatic sampler using the
equipment and methods outlined below. Minimum composite
volume must be 2 1/2 gallons. -^'
Automatic Sample Collection
Sampler - A peristaltic pump automatic sampler with ,
timer and a single glass compositing jug is required. The'2
1/2 - 3 gallon compositing bottle must be glass and cleaned
-------
- 64 -
as outlined below. New unused tubing must be used for the
sampling line and for the pump for each individual outfall or
sample location. Vacuum type automatic samplers may be used
provided that the sample chambers are glass and that they are
/ cleaned after every use as outlined for glass composite
containers. Place the sampler or composite container in an
Insulated chest and ice. Maintain the sample at 4°C during
the compositing procedure. At the completion of the
compositing period seal the container with a teflon lined
cap. Place the container in an. insulated shipping container,
1ce, and seal, then ship to the analytical laboratory.
t Maintain at 4"C during transport and storage prior to
'' analysis.
When sampling raw untreated industrial discharges which
are generally high in suspended solids it is imperative that
adequate sample flow rate be maintained throughout the sample
train in order to effectively transport the solids. In
horizontal runs, the velocity must exceed the scour velocity,
while in vertical runs the settling or the fall velocity must
be exceeded several times to assure adequate transport of
solids in the flow. The equipment used in sampling raw
discharges than must have a minimum intake velocity of 2 feet
per second. In the sampling of treated effluents just about
any commerically available automatic liquid sampler could be
used.
When more than one laboratory is involved in the
analysis of the various parameters, the sample should if at
all possible not be divided in the field but rather at the
contractors' laboratory. For purpose of this program the
composite will be divided into four parts, one part for
metals analysis, one for pesticides and PCB's, one for GC/MS
compounds and one for the classic parameters.
Blend the composite sample to provide a homogeneous
mixture including a representative suspension of any solids
in the container. No specific method is required, hand
stirring with clean glass or teflon rods, mechanical paddles
or magnetic mixing with teflon coated stirring bars may be
used. Metal mixing devices may not be used.
Metals - Withdraw a well blended aliquot of the
composite sample. Using a glass funnel, rinse the sample
container with a small portion of the sample, then transfer
-------
- 65 -
250 - 500 mcf/1 of sample to the bottle. Do not add any
preservative to the sample just seal and prepare for
Shipment. AlT~samples must be carefully identified using
labeles supplied by EGO. Indicate on the label whether the
sample is a raw discharge or treated effluent as shown. If
sample is to be run on the plasma unit only indicate so at
base of tag. Ship samples to the Chicago Regional Laboratory
at the addressed shown.
U.S. Environmental Protection Agency
Region Y, Central Regional Laboratory
1819 W. Pershing Road
Chicago, Illinois 60609
Raw discharge or treated effluent
N2 V002200
Location
Sampler
Sample Point.
Timm
rf^/
Plasma only
-------
- 66 -
Field Blank Procedure for Automatic Samplers
Blank Water - Blank v/ater must be as free from organic
Interferences as possible. The analytical laboratory should
supply this water in bulk glass containers (m'nimum of five
liters) for field use. The supplying laboratory shall analyze
the blank water to determine the organic background that may be
present.
Procedure - All parts of the sampling system must be scrubbed
with hot detergent water and thoroughly rinsed with tap v/ater and
blank v/ater prior to use. Further rinsing with methylene
chloride is required when parts permit, i.e., are"not susceptible
to dissolution by the solvent. (Note: Tygon plastic tubing is a
source of ph thai ate ester contanii nation. Where its use is
required, i.e., in the peristaltic pump, the length must be kept
as short as possible. Teflon is acceptable and may be used in
other parts of the sampling system as in intake lines. In the
field, pump two liters of blank water through the sampling line
and pump tubing and discard. Then pump three liters of blank
water through the system and collect as a blank in a 1-gallon
sample bottle that has been prepared as described below. Seal
the bottle with a Teflon lined cap. Immediately ice the blank (4°
C) and maintain at (4°C) during the transport and storage prior
to analysis.
Composite Container - Prepare narrow-mouth glass sample
bottles for use by washing with hot detergent water and
thoroughly rinsing with tap v/ater and blank water. Heat the
bottles at 4CO°C in a muffle-furnace or dry heat sterilizer for
30 minutes or alternatively, rinse with methylene chloride and
air dry .at_rqonf>empefature protected from atcrnspheric or other
sources of contamination. Caps for the bottles must be lined
with Teflon which has been solvent rinsed as above.
2. Collection of Grab Samples
Collect grab samples ( minimum of one per day) for the
analysis" of phenol, cyanide, and volatile organics
(purgable). Collect samples from the raw process discharge,
the treated effluent, and the treated effluent after
chlorination, when chlorination is practiced. It is
recommended that the samples be collected from miji'^chajinel
at nrfd-dep_th. Samples should be collected at a turbulent,
well mixed section of the channel.
-------
- 67 -
Cyanide (Total)
Container - Use nev/ one-liter plastic bottles that will
not contaminate the sample. Wash the bottles and caps with
hot water and thoroughly rinse with tap water and blank
water.
Collect a 1-Hter sample.
Pretreatment and Preservation - Oxidizing agents
such as chlorine decompose many cyanides. Therefore, at
time of collection, samples must be treated to eliminate
such agents. Test a drop of the sample at the time of
collection with potassium iodide-starch test paper
(Kl-starch paper); a blue color indicates the need for
treatment. Add ascrobic acid, a few crystals at a time,
until a drop of the sample produces no color on the indicator
paper. Then add an additional 0.6 g of ascorbic acid for
each liter of sample volume. Then add 2 ml of 10 N sodium
hydroxide per liter of sample (pH >_ 12).
Seal the sample bottle and place in an insulated chest
and tee (4*C). Seal the chest and ship to the analytical
laboratory. Maintain at 4'C during transport and'storage
keep out of direct light prior to analysis.
Phenols
Container - Use new one-liter glass bottles. Wash the bottle
and Teflon cap liner with hot water and thoroughly rinse with tap
water and blank water. 1
Collect a 1-liter sample.
Preservation - At the time of collection, acidify the sample
by addition of phospheric acid or sulfuric to pH 4. Note volume
of acid added on sample tag. Seal bottle, place in insulated -
chest and ice (4"C). Seal chest and'ship to analytical
laboratory. Maintain at 4°C during transport and storage. Keep
out of direct light prior to analysis.
-------
- 68 -
Organlcs (Purge and. Trap Method)
Containers - Use 45 to 125 ml screw cap glass vials with
Teflon faced si 1 cone septa: '
Vialsfa)- Pierce #13074 or equivalent
Septafa)- Pierce #12722 or equivalent
Wash the bottles, septa, and caps with hot water and
thoroughly rinse with tap water and blank water. Keat the
bottles and septa at 105°C for one hour, cool to room temperature
1n an enclosed contaminant free area. When cool, seal bottles
with septa (Teflon side down) and screw cap. Maintain the
bottles in this condition until just prior to fillinc with blank
water or sample.
Available from Pierce, Inc. Box 117, Rockford, IL 61105.
Collect duplicates 45-125 ml samples each time samples are
collected. Two blank water samples, sealed in 45 ml vials, are
to accompany the sample bottles curing shipment to and from the
sampling site. If preservation for residual chlorine is to be
used, collect four samples during each sampling period. Two
should be preserved and two not preserved. Two preserved and two
non-preserved blanks are to be provided.
Filling and Sealing Bottles - Slowly fill each container to
overflowing. Carefully set the container on a level surface.
Place the septum (Teflon side down) on the convex sample
meniscus. Seal the sample with the screw cap. To insure that
the sample has been properly sealed, invert the sample and
lightly tap the lid on a solid surface. The absence of entrapped
air bubbles indicates a proper seal. If air bubbles are present,
open the bottle, add additional sample, and reseal. The sample
must remain hermetically sealed until it is analyzed.
Preservation - Preservative (sodium thiosulfate or sodium
bisulfite) is used to stabilize samples containing residual
chlorine. The production of chloroform and other haloforms
continues in such samples if they are not stabilized. Waste
streams that have been treated with chlorine should be tested on
-------
- 69 -
site to determine whether or not preservative is needed. If
preservetation is required, collect both preserved and non-
preserved samples. Wrap the samples with water proof packing
material, place in an insulated chest and ice at 4°C. Maintain
at 4*C during transport and storage prior to analysis.
3. Identification of Samples
All samples and blanks must be carefully identified
using water proof labels and water proof ink. Include the
following information on the label: sample number, date and
hour of sampling, complete information as to source and
sampling point, preservative added, if any, and name of
person collecting the sample (include address and/or phone
number).
-------
APPENDIX IV
-------
REFERENCE NO. 1
DETERMINING VOLATILE ORGANICS AT MICROGRAM-PER LITER LEVELS
-------
;^V;-^;:'?^ --:^".:^'i
DECEMBER 1974
Determining
Volatile Organics
atMicrogram-
per-Litre Levels
by Gas
Chromatography
T. A. Be/far and J. J. Lichtenberg
A Metricized Article
A contribution submitted to the JOURNAL on Nov. 7, 1974, by
T. A. Bellar and J. J. Lichtenberg, res. scientists. Methods Oev.
and Qual. Assurance Res. Lab., EPA, Natl. Envir. Res. Ctr, Cin-
cinnati, Ohio.
Presented here is a method for quantitative recovery
of volatile organic compounds followed by a
description of apparatus and procedures employed to
detect 0.5 /xg/l of the substances.
Recent legislation1-2 requires analytical methods for the
determination of hydrocarbon and chlorinated organic sol-
vents in wastewater. In some cases a minimum detectable
limit of 1 /ig/I (10~3ppm) is required for specific compounds.
It is the responsibility of the EPA's Methods Dev. and Qual.
Assurance Res. Lab. to evaluate existing methods, and when
necessary, to develop new methods to meet such needs.
Determination of these substances at the 1-p.g/l (lO'-'ppm)
level has been difficult. Commonly used techniques such as
direct aqueous-injection gas chromatography, liquid-liquid
extraction, and head-gas analysis have proved inadequate.
Direct aqueous-injection gas chromatography3- * although
generally useful for analysis of industrial effluents, provides
an approximate limit of detection of only 1 000 ^ig/l (1 ppm).
Liquid-liquid extraction methods using low5 or high6 boiling
organic solvents followed by gas-chromatographic analyses
have been investigated. These methods have provided erratic
or low extraction efficiencies for volatile compounds. In addi-
tion large solvent responses and solvent impurities can cause
serious chromatographic interferences. Distillation tech-
niques7 have been employed in which a small quantity of sam
pie distillate is collected and analyzed by direct aqueous-injec-
tion gas chromatography. Detection limits of approximately I
/ig/1 (10~J ppm) have been reported for water-soluble
volatiles using this method. Poor recoveries render the meth-
T. A. BELLAR AND J. J. LICHTENBERG 739
-------
TABLE i
Trap-Saturation Volumes
Compound
Met hint
Ethane
Propane
n-Buune
a-Pentane
a-Huine
D -Aitanes
CT-CIS
Benzene
Touione
Methylene
chloride
Chloroform
Aldehydes
C2 and above
Phenols
Naphthylene
Chloro benzene
o-Dichlorooenzene
1.2.4 Tncbloro-
benzene
Silica
Gel
Layer
ml
<5'
<25'
>50'
>500'
>500-
>5W
t
>5oo-
>500*
«
<
Nonquanu-
tau e
Pore Pak
Q
ml
<5"
<5'
<50'
<100'
<250t
>500T
>500t
>500*
>500t
>55W
>500-
>500-
>500'
t
t
t
Chromo-
sorb
103
ml
<5'
500t
>500T
>500t
>soot
>50W
>SW
>500*
>soo-
>550W
>50W
>500«
Tenax
GC
ml
' T
t
<
t
t
>soot
>500T
>500T
>soot
>5005
>55W
>500*
>50M
Retenuon
Index
100
200
300
400
500
600
700-1 500
_
_
_
_
_
_
_
_
_
_
•Values reported by Bellas and Sigsby '
rvalues determined using water-saturated nitrogen as purge gas are same as those reported
under dry conditions.
• Not determined
{Determined Tor this study
TABLE 2
Purging of Selected Compounds From Water
Nitrogen Purge Gu
Percentage Remaining in Aqueous Phase
Bow Rate
mlAntit
20
20
20
20
13
13
13
13
Volume
ml
0
20
100
300
0
6J
85
143
Methylene
Chloride
100
60
0
0
100
67
30
6
2
•J0.1
Chloroform
100
55
0
0
100
94
29
0
Solubility in w
1
Boiling p
61.3
Benzene
100
46
3
0
100
71
6
0
net— per ant
0.08
»ni— C
80.1
2-Buunone
100
95
%
80
100
100
36
74
35
79.6
TABLE 3
System Response to Methylene Chloride
Concentrauon
u.X/1
5.2
104
20.8
52.0
104.0
260.0
5204)
Slope
0.0 to Data Point
78.5
76.9
78.5
77j
76J
77.9
85.7
78.0 mean
3.88 sid. dev.
Diluuon
1/100
2/100
5/100
10/100
20/100
50/100
Stock solution
od useless for water-insoluble components.
The head-gas technique* in which a sample is sealed in a
partially filled container, has been employed for many years.
Each volatile organic compound establishes an equilibrium
between the gaseous and aqueous phase. At low concentra-
tions the ratio of the concentration in the gaseous phase to the
concentration in the aqueous phase is a constant (partition
coefficient) and is unique for each organic compound. By
analyzing the gaseous phase and applying the appropriate par-
tition coefficient, one can calculate the concentration of each
organic initially present in the aqueous phase.
Of the techniques previously mentioned, the head-gas
method has the greatest potential for meeting the needs set
TABLE 4
System Response to Chloroform
Concentration
M//
6.2
12.4
24.8
62.0
124.0
310
620
Slope
0. 0 to Data Point
32J
29.7
28.1
26.8
26.8
25.4
29.9
28.4 mean
2.35 »d. dev.
Dilution
1/100
2/100
5/100
10/100
20/100
50/100
Stock solution
TABLE 5
System Response to Benzene
Concentration
Mf//
3.5
7.0
14.0
35.0
70.0
175
350
Slope
0, 0 to Data Point
220.6
219.4
214.9
215.8
207.5
196.0
232.0
215J mean
11.4 suLdev.
Dilution
1/100
27100
5/100
to/ioo
20/100
50/100
Slock solution
TABLE 6
System Response to Toluene
Concentration
u-gll
3J
70
14.0
35.0
70.0
175.0
350.0
Slope
0. 0 to Data Point
120
120
116
115
111
105
124
116 mean
6 .27 sid. dev.
Diluuon
1/100
2/100
5/100
10/100
20/100
50/100
Stock solution
TABLE 7
Purging Efficiency at 19.5 C, Percentage Recovery
Purge
Volume
mINj
0-60
60-120
120-240
240-360
360-180
480-600
600-770
770-440
840-960
960-1080
1 080-1 200
Compound and Boiling Point
n-C5
36C
100
"-C6
69C
100
n-C7
we
98
2
n-Cg
126C
90
6
3
I
n-G,
HOC
76
12
8
3
I
n-C,,
I96C
60
15
9
4
2
2
2
I
1
1
2
n-C,3-
.'.WC
44
17
13
6
3
5
3
2
2
2
2
n-C15'
270C
2
13
27
14
3
7
5
5
4
4
7
•Not 100 per cent purged using 1 560 ml N.
TABLE 8
Purging Efficiency at 65C, Percentage Recovery
Purge
Volume
ml Nj
0-60
60-120
120-240
240-360
360-480
480-600
600-::o
770-840
Compound and Boiling Point
n-Cj
J6C
100
n-C6
69C
100
n-C7
100
s&
100
n-Cj
HOC
97
3
n-Cn
/96C
76
10
6
4
2
1
1
n-Cu-
.'J-JC
66
12
6
6
4
3
2 ^
1
•Sf
27
24
• 15
11
7
6
r 6
4
'Not 100 per cent purged
740 RESEARCH
JOURNAL AWWA
-------
!)•»- l/4-.n.-OD Inltt
1/4-m.-00 Exit
10-mm Gin
Mtdium Poresitv
Simon Inltt
S-mm-OO Rubber Swtum
10-mm 00
- l/4-in.-OO Inltt
3odv t
Aavnbly —^f~
Conntctinq - m | f
Nua X^S
Stem £
Awmbly — »W
1
^ Tno V«nt
I
5
3
ii
^ cm
• Glw-Wool Plug ~Z 16
15
^ ^
— 13
— 12
50-«OMwn 11
Porous Polvmtr Sttds \Q
-^ 9
— 3
7
5 Ptf Cjnt Hiqn —
T-motraturB ciastomtr- ~— 4
200 on SO-80 Mtin _Z 3
Chromo*oro-G ~ -
— 2
3l«»-
-------
maximum bubble contact time and efficient mixing.
Gaseous volumes above the sample reservoir are kept to a
minimum to provide efficient transfer characteristics and
allow sufficient space in which most foams can disperse. Inlet
and exit ports are constructed from 0.06-mm (l/4-in.)-OD
medium- or heavy-wall tubing so that leak-free removable
connections can be made using "finger-tight" compression
fittings containing plastic ferrules. The optional foam trap is
used to control occasional samples that foam excessively.
Trap. The trap (Fig. 2) is a short section of stainless-steel
tubing packed with an adsorptive material such as gas-chro-
matographic grade porous polymers, silica gel, or molecular
sieve. Volatile materials are transported directly from the
purging device into the trap by the purge gas. The adsorbent
retards the flow of the purged compounds while the purge gas
is vented. The properties of the adsorptive material are cho-
sen to meet the needs of the particular analysis. The following
criteria must be met.
The volume of the purge gas passing through the adsorbent
packed in the trap can approach but not reach the retention
volume of the compound to be trapped (See Table I).
The retained compounds must not be irreversibly sorbed
by the trap. (Silica gel irreversibly adsorbs some aromatics
above C?).9
No chemical reactions or rearrangements may take place
when the sample is being concentrated, stored, or desorbed.
(Silica gel causes externally bonded olefins to rearrange to the
cis- and trans-2 olefins).^
The adsorptive material must be thermally stable.
Chromosorb 103 and Tenax GC have been found to perform
satisfactorily. (Divinyl benzene crosslinked porous polymers
out-gas extraneous compounds causing serious interferences
during mostgas-chromatographic analyses.)10
The trap is assembled and packed with the appropriate ad-
sorptive material according to Fig. 2. The body assembly acts
as a seal for the exit end of the trap. The modified stem assem-
bly is used to attach the trap to the desorption device. The cap
is used to seal the inlet end of the trap when it is not in use.
Desorbers. The desorbers (Fig. 3, 4) are used to transfer the
contents of the trap to the gas chromatograph for analysis.
This is done with the use of an auxiliary carrier flow-control
system which back/lushes the trap at elevated temperatures
directly onto the gas-chromatographic column. Desorber 1 is
used exclusively with one type of gas chromatograph, but
desorber 2 can be used as a universal desorber for all gas
chromatographs with a septum-type liquid-inlet system.
Desorber 1 (Fig. 3) is attached directly onto the gas-chro-
matograph liquid-inlet system after removing the septum nut,
the septum, and the internal injector parts. The modified body
assembly is screwed onto the inlet system using the TFE
gasket as a seal. A plug is attached to one of the stem assem-
blies. The assembled parts, simply called "the plug," are used
to seal the desorber whenever the trap is removed to maintain
the flow of carrier gas through the gas-chromatographic col-
umn. The flow controller, TFE tubing, and stem assembly are
used to provide the trap-backflush flow. This entire assembly
is also used to provide gas flow to operate the purging device.
Desorber 2 (Fig. 4) may be attached to any gas chroma-
tograph by piercing :he GC liquid-inlet septum with the nee-
dle. The desorber is assembled according to Fig. 4 with inter-
nal volumes and dead-volume areas held to a minimum. The
heat source is concentrated near the base of the desorber so
that the internal seals of the body assembly do not become
damaged by heat. Tine use of a detachable needle assembly
from a microsyringe makes it easy to replace plugged or dulled
needles. The flow controller, TFE tubing, and stem assembly
are used to provide the trap-backflush flow. This entire
assembly is also used to provide gas flow to operate the purg-
ing device.
A. gas chromatograph was equipped with dual-flame
ionization detectors and a microculometric detector (halide
mode).
Column 1 consisted of dual, stainless-steel, 180-cm (6-ft)
long x 2.67-mm (0.105-in) ID columns, packed with
Chromosorb-101 (60/80 mesh). The carrier gas was nitrogen
at 50 ml/min (O.cu ft/hr). The oven temperature was isother-
mal 190C (310F) or programmed from 120C to 225C (247F to
437F) at 10C(50F)/min.
Column 2 consisted of dual, stainless-steel, 91-cm (3 ft) x
1.65-mm- (0.065-in.)- ID columns packed with 4 per cent
SE-30 on Chromosorb-P (NAW) (60/80 mesh). The carrier
gas was nitrogen at 50 ml/min (0.1 cu ft/hr). The oven tem-
perature was programmed from 60C to 230C (140F to 446F)
at 10C(50F)/min.
The GC-MS system consisted of a gas chromatograph'
with a mass spectrometer! controlled by a data-acquisition
system.* The column was glass, 240-cm (8-ft) long x 2-mm
(0.078-in.) ID and packed with Chromosorb-101 (50/60
mesh). The carrier gas was helium at 30 ml/min (0.06 cu ft/
hr). The initial oven temperature was 125C (257F) for 3 min
and then programmed at 4C (39F)/min to 220C (428F).
Reagents
Organic-free water was prepared by passing distilled water
through a water-treatment system.§
Standard stock solutions were prepared by injecting 1-5 p.\
61.02 x lO^cu. in. of the compound to be determined into a
1-1 (61-cu in.) volumetric flask partially filled with organic-
free water. The mixture was then diluted to volume with
organic-free water to give concentrations between 1 and 7
mg/1 (1 and 7 ppm). Dilutions were made from the stock solu-
tion by pipetting a known quantity of stock solution into a
partially filled volumetric flask and diluting to volume with
organic-free water. [For low-level work (1-10 /ig/0
(10~3-10'2ppm) a 1:10 dilution of the stock solution was pre-
pared, and secondary dilutions were prepared.]
Procedure
Purging and trapping. With nitrogen gas flowing through the
purging device (Fig. 1) at 20 ml/min (0.04 cu ft/hr), the trap
inlet (Fig. 2) was attached (finger-tight) directly to the purg-
ing device exit using a compression fitting. The trap vent was
inserted into the exit end of the trap. Five millilitres of sample
were injected into the purging device and purged for the
specified time (11 min). The trap was then removed from the
purging device, and the vent plug was removed and replaced
with a cap to seal the trap inlet.
Trap conditioning. Newly packed traps were conditioned at
approximately 200C (392F) with a nitrogen flow of 20, ml/min
(0.04 cu ft/hr) for 16-24 hr with one of the desorbers and
vented to the room. Prior to daily use, traps were placed into
the desorber and conditioned at 130C (266F) for approxi-
mately 10 min while being backflushed with nitrogen at 20
ml/min (0.04 cu ft/hr).
Dasorption and analysis. Desorber 1 (Fig. 3). The gas-cjjroma-
•Varian Aerogripn See. 1400
tFinnigan 1015C Quadrupole
tSysienu Industries 150
§M>IIipore Super-Q
742 RESEARCH
JOURNAL AWWA
-------
tographic oven was cooled to below 30C (86F) by leaving the
oven door open. The plug was removed from the desorber,
and the cap was removed from the trap; the trap was then in-
serted into the desorber and locked into place. The trap-back-
flush flow fitting was then locked into place on the trap-exit
flow fitting and backflushed with nitrogen at 20 ml/min (0.04
cu ft/hr) for 3 min between 125C and 130C (257F and 266F).
The trap-backflush flow fitting was removed with the trap still
locked into place, the oven lid was closed, and the oven was
rapidly heated to its normal or initial operating temperature. A
gas-chromatographic analysis was carried out under these
conditions.
After analysis the trap was removed by: (1) inserting the
trap vent into the trap-exit fitting (to vent-inlet system), (2)
removing the trap, (3) resealing GC-inlet system with the
plug, (4) removing the trap vent, and (5) resealing the trap in-
let with the cap.
Desorber 2 (Fig. 4). The gas-chromatographic oven was
cooled to below 30C (86 F). The needle was inserted into the
liquid-inlet system on the gas chromatograph. The trap was
then inserted into the desorber and locked into place. The
trap-backilush flow fitting was locked into the trap-exit flow
fitting and backflushed with nitrogen at 20 ml/min (0.04 cu ft/
hr) for 3 min between 125C and 130C (257F and 266F). After
desorption and sample transfer were completed, the needle
was removed from the liquid-inlet system, the oven lid was
closed, and the oven was rapidly heated to the initial operating
temperature. Gas-chromatographic analyses were performed
under these conditions. After sample transfer the trap was
removed and sealed for future use.
Investigation of Method Parameters
Initial studies were carried out to determine the volume of
purge gas needed for quantitatively extracting selected
volatile materials from a water sample. The purging device
was charged with 5.0 ml (0.3 cu in.) of an aqueous solution
containing methylene chloride, chloroform, benzene, and 2-
butanone concentrations, each in excess of 10 mg/1 (10 ppm).
As the solution was being purged with nitrogen, 3-^1 ali-
quots were periodically withdrawn for analysis by direct-
aqueous injection. Analyses were performed on the aqueous
mixture until the concentrations of the dosed materials were
reduced to or below the limit of detection, approximately 100
Mg/1 (10~l ppm). This experiment was initially performed
with a purge-gas-flow rate of 20-ml/min (0.04-cu ft/hr) nitro-
gen. The ilow rate was reduced 65 per cent to 13 ml/min (0.03
cu ft/hr), and the experiment repeated. The percentages of
the dosed compounds remaining in the aqueous phase with
respect to the purge volume are listed in Table 2.
Those trap saturation volumes reported in Table 1, desig-
nated by footnote t, were obtained by Bellar and Sigsby9 for a
dry-air system. To determine what effect, if any, water that is
inherent to the system reported herein, would have on the
saturation volumes, the authors redetermined the volumes
using water-saturated nitrogen as the purge gas; little if any.
change was observed. The saturation volumes for several
organochlorme compounds, not prsviously reported, were
also determined under this condition.
The purging-and-trapping system was tested with selected
industrial solvents over a wide range of concentrations.
Ideally the response for each compound would be linear over
the entire concentration range. By using the standard solu-
tions and operating parameters previously described, the
autho'rs obtained the data listed in Tables 3-6. The peak
DECEMBER 1974
height of each compound was measured and divided by the
concentration to give the slope between 0,0 and each data
point collected. Response curves for four common organic
solvents are shown in Fig. 5. The standard deviations from the
mean slope are also listed in the tables.
To determine the effect of variation in the physical proper-
ties of individual compounds on the efficiency of the system,
the authors tested a homologous series of n-alkanes. The test
mixture consisted of n-C5 to n-C15 in organic-free water. This
mixture was analyzed according to the prescribed procedure
using a Tenax trap. Tenax was used as the adsorbent because
it has a higher thermal stability than Chromosorb 101 and can
be operated at the temperatures required for desorbing the
higher molecular-weight alkanes. To determine the purge
volume required for quantitative transfer of hydrocarbons
over the wide boiling range, successive fractions were col-
lected at ambient temperature (19.5O [67F) and analyzed by
flame ionization (FID) gas chromatography using an SE-30
column (See Table 7). The test was repeated at an elevated
purging temperature (65C) [149F] (Table 8).
When the method was applied to a sample from a local sew-
age plant which serves a diverse industrial area, the compli-
cated FID gas chromatogram shown in Fig. 6 resulted. The
sample was analyzed again using the microcoulometric detec-
tor which gave the chromatogram shown in Fig. 7. The com-
pounds identified in the chromatograms were confirmed by
GC-MS.
Results and Discussion
The data in Table 2 show that it is possible to purge the
water insoluble (<2 per cent soluble) compounds from 5 ml of
water using < 150 ml (9 cu in.) of nitrogen. A decrease in the
purge-gas flow rate of 65 per cent indicated that a slight in-
crease in the volume of purge gas is needed for quantitative
transfer. Water-soluble materials whose partition coefficients
do not favor the gaseous phase are only qualitatively trans-
ferred regardless of the purge volume.
Trapping. Judging from the data reported by Bellar and
Sigsby9 and other data exhibited in Table 1, one can see that
organics contained in small volumes of water-saturated nitro-
gen can be concentrated. It is apparent from these data that
compounds with a retention index > 500 can be quantitatively
purged and trapped. Retention indices given in the literature
on porous polymers11'13 make it possible to predict trap satura-
tion volumes for a wide variety of organic compounds. Since
most hydrocarbons and substituted hydrocarbons commonly
present in wastewaters have retention indices >500, porous
polymers were used in developing this method.
Water has a retention index of < 300 and is not quan-
titatively trapped by porous polymers. Therefore, gas-chro-
matographic columns and detectors adversely affected by
water can be used with a minimum of interference.
The statistical data generated in Tables 3-6 reflect an ac-
cumulation of errors for the entire method. After one con-
siders the number of manipulations involved and that gas-
chromatographic errors are generally ±3 per cent, it appears
that this is, indeed, a useful method. Further study of these
data indicates that the majority of the errors are caused by the
volumetric-dilution procedure. The larger the pipet used to
withdraw aliquots of the stock solution, the larger the error. A
buret may be a more suitable device for delivering volatile
solutes.
For the compounds studied, based on data in tables 2-6, the
authors estimated that purging transferred at least ()')
T. A. BELLAR AND J. J. LICHTENBERG 743
-------
P - cent of the volatile, water-insoluble compounds from the
aqueous phase to the gaseous phase. The data in Tables 3-6
a;id some unreported duplicate data show that the purging
etfiaency is identical from 2 500 ^g/1 (2 500 x 10~3 ppm) to at
kast 6 /Ag/1 (6 x 10~3 ppm). Therefore, the compounds
studied can be quantitatively determined over that concentra-
tion range.
Further study of the data in Table 7 indicates that the
aikanes up to €9 can be quantitatively purged using < 500 ml
<30 cu in.) of purge gas. Purge volumes exceeding 1.5 1 (91.5
c- in.) failed to transfer 100 per cent of the Cn through C}S
aikanes. Raising the temperature of the purging device and
sample (Table 8) extended the useful range of the method to
C,, hydrocarbons. If a water sample contains volau'les over
the entire boiling range represented by these data, it may be
necessary to trap two fractions in order to perform a complete
quantitative analyses on the sample. This is apparent from the
data in Table 1 that show that compounds with a retention in-
dex of < 600 will saturate the trap and be vented before the
high boiling materials are quantitatively purged.
Swnp<« preservation. Because of the volatility of the organic
materials detected by this method, common sample-preserva-
tion techniques are inadequate.3-'* The simplicity of the trap-
and-purging device makes it possible, however, to collect,
purge, and trap the sample at the sampling site. The trap and
contents can then be sealed and shipped to the laboratory for
analysis, and thus, the need for sample preservation is elimi-
nated.
Application of the method. Judging from the experimental
data reported in this article, one may see that this method has
great potential for the analysis of trace-volatile organics con-
tained in a wide variety of water sources. For quantitative
determinations the method is limited to organic compounds
that are < 2 per cent soluble in water and boil below 200C
(392F). Significant qualitative enhancement of compounds
whose boiling points exceed 200C (392F) can be expected
wnen the sample is heated. The method is useful from 1 to
2 500 jig/1 (10~3 to 2 500 x 10~3 ppm) with the use of most gas
ctromatographs. At concentrations exceeding 2 500 /ig/1
(2 500 x 10~3ppm) chromatograph-column flooding and non-
linear-detector responses generally occur. Since direct-
aqueous injection techniques are useful down to 1 000 /xg/1
(\ 000 x 10~3 ppm) the two methods can be usedtogether to
perform analyses over a wide range of concentrations. For
water-soluble compounds the distillation technique should
provide the supplemental methbdology needed to analyze
most industrial effluents and natural waters.
\ wide variety of wastewater samples were analyzed using
ti;-; described method. The chromatograms (Fig. 6, 7) show
the results of one such analysis. Qualitative identifications
^ere made using desorber 2 and a GC-MS system.* The
quantitative analyses were obtained using desorber 1 with a
microcoulometric detector. Only one of the peaks in the FID
chromatogram have been identified. At the sensitivity ranges
s.nown, only the chlorobenzenes are likely to appear on the
FI~' chromatogram.
The method worked well except for the following: one sam-
ple collected from a sewage-treatment plant foamed ex-
:assively and caused water to be transported from the purging
Jev.c* into the trap. Decreasing the sample size from 5 to 3 ml
(0.3 to 0.2 cu in.) or using the foam trap eliminated this prob-
lem. Water entering the trap causes nonquantitative trapping
*Snni|*n System
744 RESEARCH
Reprinted .ir.i -orTl^ced as .1 part 3:
Journal'American Wjcer Works Assn.
Vol. 66 NJ. 12 Deceober [97i
?r'..-.:ii In U.S.A.
and severe gas-chromatographic interferences.
When water samples contained gross amounts of water-sol-
uble organics, a sufficient quantity of these materials was col-
lected in the trap for detection. When only water-insoluble
materials were present in the sample, it was found that the
purged water could be withdrawn with a syringe and the purg-
ing device could be recharged for successive analyses. When
large concentrations of water-soluble organics were present, it
was necessary to dry the purging device in an oven at HOC
(230F) before an interference-free successive analysis could
be performed. Other researchers14-l5 have reported on similar
methods for the analysis of aqueous samples; their work has
been primarily qualitative.
This current work has shown that the method can be used
for the quantitative measurement of a wide variety of water-
insoluble compounds whose boiling points are < 150C (302F).
By slightly modifying the method, one can also quantitatively
measure materials that boil at approximately 200C. Qualita-
tive sample concentration occurs for a wide variety of other
materials for which quantitative measurements could possibly
be made if recovery factors were experimentally determined.
Vinyl chloride is one compound of considerable interest that
can be determined by this method. Analytical conditions for
this specific application are under investigation.
Summary
The method for quantitative recovery and gas-chroraa-
tographic determination of water-insoluble, volatile organic
compounds presented here provides a detection limit of ap-
proximately 0.5 Mg/1 for many compounds.
References
1. National Pollutant Discharge Elimination System, Proposed
Forms and Guidelines for Information from Owners and Opera-
tors of Point Sources, Pt. 2. Fed Rgtr.. 38:75:9783 (Apr. 19,1973).
2. Ocean Dumping Criteria, Pt 2. Fed. Rgtr., 38:94:12872 (May 16,
1973).
3. SUGAR, J.W. & CONWAY, R. A. Jour. WPCF, 40:9:1622 (Sep.
1968).
4. Tentative Recommended Practice for Measuring Volatile
Organic Matter in Water by Aqueous-Injection Gas Chroma-
tography, Annual Book of ASTM Standards, Pt. 23, Water.
ASTM D 2908-70T, Atmospheric Analysis (1973).
5. Methods for Organic Pesticides in Water and Wastewater. EPA,
Natl. Envir. Res. Ctr., Cincinnati, Ohio (1971).
6. DUENBOSTEL, B.F. Method for Obtaining GC/MS Data of Volatile
Organics in Water Samples. Internal Rprt. EPA, Region II,
Edison, N.J. (May 14, 1973).
7. Procedure for Water Soluble Volatile Organic Solvents in
Effluents and Streams. Org. Lab., Chern. Svces. Br., Region 4,
EPA. Athens, Ga. (Aug. 1973). '
8. Chlorinated Organics and Hydrocarbons in Water by Vapor
Phase Partitioning and Gas Chromatographic Analysis. Method
No. QA-466, Dow Chemical, Louisiana Div., Plaquemine, La.
(Jan. 1972).
9. BELLAR, T.A. & SIGSBY, J.E. Non-Cryogenic Trapping Techni-
ques for Gas Chromatography, Internal Rprt. EPA, Div. of
Chem. and Phys., Research Triangle Pk., N.C. (1970). '
10. BELUAR, T.A. & SIGSBY, J.E. The Analysis of Light Aromatic Car-
bonyls. Phenols, and Methyl Napthylenes in Automotive Emis-
sions by Gas Chromatography, Internal Rprt. EPA, Div. of
Chem. and Phys., Research Triangle Pk., N.C. (1970).
11. Chromosorb Century Ser. Bull., Johns-Manville, Celite Div.,
Greenwood Pla2a, Denver, Colo. (Nov. 1970).
12. Tenax-GC Bull. No. 24, Appl. Sci. Lab., Inc. State College, Pa.
13. HOLLIS, O.L. & HAYES, W.V. Jour. Gas Chromatog., 4:7:235 (Jul.
1966).
14. ZLATKIS, A. & LIEBICH. H.M. Profile of Volatile Metabolites in
Human Urine. Clin. Chem.. 17:7:592 (Jul. 1971).
15. NOVAK, J., ET AI_ Analysis of Organic Constituents Present in
Drinking Water. Jour. Chromatog, 76:1:45 (Feb. 1973).
JOURNAL AWWA
-------
NO. 2
CE COMPOUND TO CALIBRATE ION ABUNDANCE MEASUREMENTS
IN GAS CHRCMATCGRAPHY--MASS SPECTROMETRY SYSTEMS
-------
Copyright 1976 by
Reprinted from ANALYTICAL CHKMISTItY. Vol. 47.1'ncn W-O. June 197.r>
)76 by the American Chemjcnl Society nn
make empirical identifications of compounds in environ-
mental, hiontcdical, nnd other type* of samples. Clenrly,
correct identifications require some consistency between
reference spectra nnd observed spectra, and better quality
abundance data would improve the effectiveness of all em-
pirical search systems.
1 Author to whom correspondence should h« addressed.
In addition to an ion abun "ance calibrant, there is a need
for a reference compound to evaluate the overall perfor-
mance of a computerized GC/MS system. \Ve have ob-
served spectra with acceptable ion abundances but, be-
cause of poor resolution adjustment, broad peaks that were
interpreted by the data system as multiplets. A reference
procedure would allow an operator to validate the perfor-
mance of the GC column, the sample enrichment device,
the ion source, the ion detection circuits, the analog-to-dig-
ital converter, the data reduction system, and the data out-
put system. The application of this procedure would en-
hance the overall quality of results emerging from the sys-
tems in use.
There is a special need to closely monitor the perfor-
mance of the RF quadrupole mass spectrometer. Unlike
the magnetic deflection spectrometer, the active ion sepa-
rating device of the RF field spectrometer, the rods, is di-
rectly contaminated during operation. After prolonged op-
eration, the rods are subject to severely degraded perfor-
mance which usually affects the region above 300 amu first.
Often this degraded performance is not detected because
there is no generally accepted performance standard to
form the basis for such judgments.
The Environmental Protection Agency has developed
and used experimentally a performance evaluation/abun-
dance calibration procedure for the last several years. A set
of chemical and physical properties criteria for a reference
material was developed and a number of likely candidates,
including PFK and PFTBA, were tested. The compound
decafluorotriphenylphosphine (DFTPP) was selected as
the one which met most of the criteria.
This paper reports the criteria on which t.he compound
was selected, its mass spectrum, some physical and chemi-
cal properties, and some performance data that were col-
lected over the last few years. An RF field mass spectrome-
ter, which has been tuned to give the suggested inn abun-
dances in the reference compound spectrum will, in gener-
al, generate mass spectra of organic compounds which nre
very similar to spectra generated by other types of mass
spectrometers. Thus RF Held mass spectra become directly
comparable to spectra of compounds in collections which
have been obtained with otjier types of mass spectrome-
ters.
EXPERIMENTAL
Material*. All chemicals unii solvents wore obtained from ofim-
mcrcinl suurcet. DefnlluorolriplicnylpluiNphiiiti woa prrjinrcd ac-
ANALYTICAL CHEMISTRY. VOL. 47. NO. 7, JUNE 1975 • 093
-------
conlinq to the procedure of Dun (/); m.p. (<'i -M"; Anal. Cnlrd f(ir
CiHHr.K,,,!': C, .1H.KO; II, 1.1.1. Kound: C. -trU)'., 49.0.1; H. 1.10. I 07.
Purity, lu'xcd on flame iiiniy.nlinri detector t"H chruniHl> interfaced to lh« spectrometer by an all-
' glass jet type cnricliment device and an all-i;l;is.s transfer line. Con-
trol of the quudrupolc rod ma'ss .set voltages, data acquisition, data
reduction, and data output was accomplished with a System in-
dustries data system which employed a Digital Equipment Corpo-
ration PDP-8/E mini-computer and a l.C million word Diablo disk
drive.
All of tl-c systems referred to in Table III also used this spec-
trometer and data system which has a user option to integrate ion
currents at one or more- (maximum of ten) 0.1-amu intervals be-
tween each integer mass. The maximum ion current value is .select-
ed for each amu by the control program and abundances of non-
integer ionic masses are measured correctly. With DFTPP, this op-
tion was not used since the ions in the spectrum of DFTPP have
masses very close to the integer values (e.g., M* = 441.997).
The gas chromatograms and mass spectra were displayed on a
Tektronix Model 4010 cathode ray tube or a Houston Instruments
model DP-1 flatbed plotter. .
Gas Chroraatography. Most of the work reported in this paper
was carried out using a 6-ft X 2-mm (i.d.) glass column parked
witK 1.95% QF-1 plus 1.5% 0V-17 on 80/100 mesh Supelcoport.
The flow rate was about 30 ml/min; column temperature, 180°; in-
jector temperature, 210°; and interface oven-transfer line tempera-
ture, 200-210°. The compound decafluorotriphenylphosphine was
also chromatographed on a variety of other columns of van-ing
length and stationary phases. In general ihese were 4-8 ft, metal or
glass, 100-250° column temperatures, and 20-35 ml/min flow
rates. Stationary phases were 3% SE30, 5.5% OVl. 3-5% OV17,
2-6% OV101, Dexil 300, and 0.1% OV210.
Chromatography was also successful on a 100-ft, 0.02-in. (i.d.)
support coated open tubular column coated with QF-1. In general,
• retention times of 4-10 minutes were observed. Cross-linked po-
rous polymer packed columns were not suitable for this compound.
Similarly, a 7-ft coiled glai>s column (i.d. 2 mm) packed with 10%
free fatty acid phase on 60/80 mesh chromosorb \V gave poor re-
sults.
Procedure. A stock solution of DFTPP at 1 mg/ml (1000 ppm)
concentration in acetone was prepared. This stock solution was
shown, by repeated analyses, to be 97%+ stable after six months.
and indications are it will remain usable for several years. An ali-
quot of the stock solution was'diluted to 10 up/ml (10 ppm) in ace-
tone. The very small quantity of material present in this very di-
lute solution is subject to depreciation because of adsorption on
the walls of the glass container, reaction with trace impurities in
the acetone, etc. Therefore the dilute solution was used fur only a
short term, i.e., 1-2 weeks.
The gas chromatographic operating parameters were adjusted to
permit the acquisition of at least four complete mass spectra dur-
ing the clution of the DKTPP. The mass/charge scale of the mass
spectrometer was calibrated according to ihe standard procedure
provided by the manufacturer. The muss spectrometer and data
system were prepared for CC data acquisition using the following
parameters: mass ran^c. 3.1-500 amu; electron energy, 70 eV; trap
current. 250-500 »«A; preamplifier sensitivity, 10~7 A/volt; electron
multiplier voltage, 3000 volts; and muss spectrometer manifold
temperature, 100°. Under these conditions, the ion source temper-
ature of the Finnigan mass spectrometer in not known. The pres-
sure in the spectrometer WHS about 1<>~' Torr nnd the base line was
adjusted with the automatic zero program. The spectrometer dula
system was set to integrate the preamplifier signal for 8 msec at
each integer ma.is unit. Alternatively, the iii(ej;nilion time as a
function of si|;nnl strength option was iililixrd. This will \tf de-
scribed in detail in a future publication (2).
An injection of JO n;; CJ jj) of the dilute standard wan mndc nnd
datn Acquisition WHS In-guii iiltcr most of I lie solvent WH.H pumped
from (he spectrometer. Duiii 4ic(|in.silinn wiis rum-hided niter edi-
tion of the DrTIT. The m.iiw s|*clriiin of 1)1*1*1'!' was oliliiiiicd
by KelcvtiiiK » r-pei-trniii nimilxT on the front side of the C>C, |»enk
•3 nenr tin' upvx n.s possible. A lmrk|;rniiiul s|XTlruin WHS M-leclcd
from ono of the spectra immediately (irecfilnn; the OFTI'I' |H-uk.
Severul uprrtra were sometimes plotted in un attempt to find one
which Hi the nlxindfinrc rrilcri.-i If no upertMim could l»o oliliuned
which fit the criterin, the n«l nnd ion source potent i.ilt were nd-
justed ns in the in;iniifnclurcr'» lune-up nriiwdiiri'. ll llu< failed to
priKlucc Ihe correct spectrum, more extensive mmnU-nance «:IH
performed. This was immlly cltiininu the ion source nnd/or the
qundrupolc rods. These measures usually corrected the m.-ilcondi-
lion and a spectrum of DFTl'l' could be obtained which fit the cri-
teria.
RESULTS AND DISCUSSION
The results of several recent studies illustrate the need
for a standard relative abundance calibration procedure
and pefonnance evaluation standard. A study was reported
in 1973 (3) of calibration data from various types of mass.
spectrometers. Relative abundance data were reported for
an aliphatic hydrocarbon, n-hexadccane, and an alkyiated
aromatic hydrocarbon, 1-phcnyl undecanc. The participat-
ing laboratories introduced these samples with convention-
al batch inlet systems into a variety of single and double fo-
cusing magnetic deflection and several RK quadrupole
spectrometers. Selected data from that study are given in
Table I. Measurements at the selected ions agree reason-
ably well below about mas* 100. Above mass 100, there is a
clear indication of reduced sensitivity with the quadrupole
spectrometers. This trend supports the widespread idea
that quadrupole spectrometers are significantly less sensi-
tive than magnetic deflection spectrometers at the higher
masses. The data above mass 100 obtained with the 21-491
and MS-902 spectrometers reveal the well known fact that
magnetic deflection spectrometers are susceptible to re-
duced high mass sensitivity also. This may be due to em-
phasis on low mass sensitivity during ion source tuning or
performance degradation due to contamination of the ion
source.
In late 1972, samples of DFTPP were sent by us to a
number of EPA and other laboratories. This survey was
conducted to obtain relative abundance comparisons up to
' mass 450. In addition, it was requested that the sample be
introduced with a GC inlet system and any GC column that
was convenient for the participating laboratory. The re-
sults from magnetic deflection systems are shown in Table
II and from RF quadrupole systems in Table III. The ions
selected for comparison are spaced at approximately 75
amu intervals up to mass 275 and include, in addition, the
molecular ion (M*) at mass 442 and the molecular ion con-
taining a single 13C atom at mass 443. The theoretical 443/
442 percentage is 19.8%.
Relative abundance data for DFTPP from three of the
magnetic sector instruments is in very good agreement and
all four magnetic instruments produced acceptable values
for the (M+ + 1)/M+ percentage. The relative abundance
data from the 21-490 may be an example of ion source tun-
ing to emphasize the molecular ion region or perhaps it
merely reflects the selection of a spectrum number too
close to the front of the peak. In the latter event, the mo-
lecular ion would have been observed after the concentra-
tion of the DFTPP in the ion source had increased signifi-
cantly.
The relative abundance data for DFTPP from the RF
quadrupole spectrometers were much less consistent. Lab-
oratory No. 1 reported the base peak as mass 51, laborato-
ries 2-7 reported the base peak as mass 198, laboratory 8
reported the molecular ion ns the base peak, and laborato-
ries 9-11 found muss 09 (CK.!*) n.s the base ponk. The range
of abundance measurements at nny RI'VIMI muss was pencr-
nlly mudi larger with I lip RF qundrupolc spectrometers.
For oxiintple, the three magnetic deflection spectrometers
that nu'astircd mass 198 us I ho base peak had n range of 21
relative abundance units at mass 51. The six quadrupolcs
that measured mass 198 as the base penk had a range of 39
•98 • ANALYTCAL CHEMISTRY. VOL. 47. NO. 7, JUNE 1975
-------
Table I. Selected Relative Abundance Dulu for llexudccanc Measured with a Variety of Mass Spectrometers'1
Kclatlv« ibuuJanc*,*
M«u
57
71
85
99
113
127
141
155
226
• Dala taken from Reference 3. * Sinplc focusing sector magnetic deflection spectrometer. r Double focusing (electrostatic ami magnetic
fields) modified Nier-Johnson spectrometer. '•Double focusing Maliauch-Herzog geometry spectrometer. 'Radio frequency t|ii;idrupole
spectrometer.
CH-7*
100
60
37
12
7
5
4
4
8
HMU-6»
100
55
40
12
8
7
5
5
12
21-4-W*
100
73
40
14
9
7
G
5
11
21-49K
100
65
45
12
7
5
3
3
3
21-4'jf
100
75
52
16
9
6
6
6
11
21 -non'
100
72
48
11
7
6
5
5
9
MS-DOS*
100
66
37
10
6
4
2
2
3
1015'
100
54
32
7
3
2
1
1
1
101 5"
100
60
35
8
4
2
1
1
1
Table II. Selected Relative Abundance Data for DFTPP Measured with Single
Focusing Magnetic Deflection Spectrometers and GC Inlet Systems
Percent relative abuadaacc at majs-
Sp«ctToro«l*r
Varian CH-7
Varian CH-5
Nuclide 1290G
DuPont 21-490
51
40
60
34
12
127
42
52
37
13
198
100
100
100
34
275
26
24
29
11
442
92
95
86
100
•U3
20
19
17
21
2
443/442' 10
21.7
20.0
19.8
21.0
Table III. Selected Relative Abundance Data for DFTPP Measured with
Fianigan 1015 RF Quadrupole Spectrometers and GC Inlet Systems
Percent relative abundance at ma«
Lab
1
2
3
4
5
6
7
8
9
10
11.
SI
100
81
53
53
92
86
57
14
66
93
97
127
49
50
68
48
55
40
43
19
80
57
85
193
98
100
100
100
100
100
100
42
76
85
65
275
20
13
24 .
19
22
28
16
13
19
11
11
442
51
33
31
64
57
56
48
100
47
20
2.5
443
9
7.5
5.5
12
12
10
10
91
13
4
2.5
443/442 * 102
17.6
22.7
17.7
18.8
21.0
17.9
20.8
91.0
27.7 .
20.0
100.0
units at mass 51. All four magnetic deflection spectrome-
ters produced molecular ion measurements between 86-
100%; the quadrupolc values for the molecular ion ranged
between 2.f>-100%. The values of the (M* + 1)/M+ per-
centage from the four magnetic deflection spectrometers
had a standard deviation ot'O.S"... The same values from the
quadrupole spectrometers had a standard deviation of 3%
after rejection of the 91% and 100% observations.
The more diffuse nature of the RF quadrupole abun-
dance measurements was probably due to a variety of caus-
es including the presence of generally less experienced op-
erators, the failure of some operators to utilize ion abun-
dance calibration procedures, inadequate ion source or
quadrupolc rod maintenance, more difficult qundrupole
tune-up adjuMments, and the selectipn of spectrum num-
bers too close to the front or apex of a (1C peak.
The hcxadecane spectrum was measured in this labora-
tory with on RF iniaclrnpolc niter the spectrometer was ad-
justed to give a DFTPP spectrum similar to that produced
by I he Vurinn and Nuclide magnetic deflect inn spectrome-
ters. The musses nnd relative abundances that correspond
to those in Table I were: 57. 100; 71, U5; 85, ;!!); Oil, Kl; 113
1J-_>lJ, 10.
\Ve concluded that the RF quadrupole spectrometer
could be maintained, without unreasonable effort, in a con-
dition that would produce mass spectrometric fragmenta-
tion patterns that were very similar to patterns produced
by other types of spectrometers. However, it was also clear
that a standard relative abundance calibration procedure
and performance evaluation standard was required and
that use of this standard would benefit users of magnetic
deflection spectrometers also.
Criteria for the Ideal Reference Compound. The
ideal reference compound should possess a number of im-
portant properties. It should be available in very pure form
as a crystalline solid. This is necessary to facilitate accurate
weighing and the- preparation of standard solutions to eval-
uate GC/MS system sensitivity in terms of signal to noise
for a given quantity. The compound should have high ki-
netic and thermixlynamic stability nnd be soluble in a vari-
ety of common organic solvents to facilitate gus chromatog-
rnphy. The material should be very easy to gas chrumato-
grapb on a wide variety of columns of differing polarity.
This property would encourage its application on whatever
column was of particular importance in a given laboratory.
The muss spectrum of the compound must display an
ANALYTICAL CUCMISTRY. VOL. 47. NO. 7. JUNE 1975 • 997
-------
ebundnnt molcculnr or fragment i«n ncnr mnss .r>00. This is
'an extremely important factor since mnny comjH>unds of
environmental nnd biomedical significance hiivc ions in the
400-500 umu range. The ion must be very abundant in
i puKoat
og
9
I
8
8
D
$5
$
.8
12
8
8
- .8
, .8
u.
Q
w
I
.P
.8
a
- 2
•
order to cnsily evaluate the system acnsitivitv and resolu-
tion in the high muss region. The commonly use<) mass cali-
bration compounds PFK and IM-'THA have ions HI this re-
gion, but these nre very inadequate because of their very
low relative abundance. For example, a 100% reduction in
spectrometer sensitivity at mass 500 is reflected in the
spectrum of PFTBA by a change in relative abundance of
the mass 502 ion from 2% to 1%. In contrast, the spectrum
of the reference standard should not be dominated by a
single very abundant ion which tends to saturate the detec-
tor and reduce all other ions to very small relative abun-
dances. Most fluorinated aliphatic compounds, e.g.,
PFTBA and PFK constituents, suffer from the dominance
of the mass 69 CV-.\+ ion. An even distribution of ions of
even relative abundances over a wide mass rnn^e is most
desirable. On the other hand, the compound should not
possess too many ions which might cloud a spectrum with
too much information to allow a fast evaluation of the sys-
tem performance.
It was clear that the fluorinated aliphatic compounds in-
cluding PFK and PFTBA were not suitable because of sev-
eral serious limitations. n-Hi-- tdecane is widely used as the
standard of reference in hydrocarbon type analyses in the
petroleum industry, but its low molecular weight. 226, and
the generally low relative abundance of the molecular ions
of aliphatic hydrocarbons rule out this type of ?tandard.
Cholesterol is a crystalline compound of reasonable molec-
ular weight but it is difficult to chromatograph without
derivatization. Methyl stearate is often used as a test com-
pound but it is unacceptable because of its molecular
weight, 298, and the low relative abundance of the molecu-
lar ion. Perfluorodecalin was recently proposed (•»') as a
mass calibration standard for low resolution mass spectra.
One of its attributes is that the relative contribution of
mass 69 to the total ionization is much less than for other
fluorinated aliphatics. Nevertheless, the compound is a vol-
atile liquid with no ion of greater than 10°o relative abun-
dance above mass 293. Perfluoroalkyl-s-triazines and relat-
ed compounds (5) have been used to excellent advantage as
very high mass calibration standards for the mass to charge
scale. They suffer similar disadvantages of dominance by
mass 69 and large gaps where no abundant ion is observed.
Triphenylnaphthalene (6) was reported as a useful mass to
charge scale calibrant. This compound has a molecular
weight of 356 and produces a large number of ions includ-
ing several abundant clusters.
" The compound bis(perfluorophenyl)phenylphosphine 1
(or decafluorotriphenylphosphine. DFTPP) was one of a
number of compounds evaluated as a possible iojn abun-
dance calibration reference compound and standard for
performance measurements. Its spectrum is shown in Fig-
ure 1. The compound meets nearly all of the criteria de-
scribed previously. Its spectrum contains relatively abun-
dant ions at about 75-ainu intervals (Tables II and III) be-
tween masses 51 and 275. It is deficient in that there is no
abundant ion in its spectrum between mass 275 and 442.
The molecular ion at mass 4-12 is very abundant but does
not dominate, mid there are not too many ions that would
preclude rapid inspection and evaluation of a spectrometer
performance.
no*
Ai VTW*»«I <*uca«CTOV \/ni tt MA 7
1O7S
-------
Proposed Compositions of Ions in I lie Spectrum of
DFTPP. Kx.id inns*; niiMVirr.-in<»iil.s nn.ri i-nrrcsprinds to loss of a phenyl group.
The ion ;it mass 27~> has the composition (C,;H.-,)(CfiFfi)l>'1'
which results from Id-.- loss of a single perfluoropbenyl
group from the molcr;i!;..- ion. We propose the fragmenta-
tion process in whirh this ion either loses its phenyl group
to form the mass I9S ion, or loses hydrogen fluoride to form
the tetrafliioroph(is|iha?')li: ion of mass 2f>.'j.
Mass 127 is perhaps the phenylfluorophosphine ion
CgHsP+F. The ions of m.vses .77, 69, and 51 are well estab-
lished as the phenyl, CF;+, and C4H;t+ ions. The latter is a
decomposition product of the phenyl ion and the CFn4 ion
is produced by extensive rearrangement of a perfluoro-
phenyl ion.
Relative Ion Abundance Criteria. It was our goal to
arrive at a set of relative abundances for DFTPP that
would be a standard for performance evaluations and a
guide for ion abundance calibration. The data collected in
the 1972 survey (Tablf- II and III) as well as hundreds of
repeated measurements in this and several other EPA labo-
ratories were the basis for these criteria. It must be empha-
" sized that the data from the 1972 survey were taken direct-
ly from the computer program generated plots or digital
printed data when available and that the criteria are in-
tended to apply to the same output. The data handling sys-
tem of a modern GC/VuS is an integrated part of the total
system, and the data system performance must be included
in the overall evaluation. Clearly, the computer generated
output is the most convenient for the operator to use in the
evaluation.
The majority of measurements found mass 198 as the
base peak and this vv.ns selected as the basic criterion
(Table IV). All other criteria were developed using only
those spectra which h;\d mass 198 ns the base peak. Abun-
dant ions were located at approximately 75 amu intervals
above ami below mass 1!>S. These wore masses 51, 127, and
275 nnd they were included in the criteria to provide a men-
sure of system sensitivity ut regular intervals throughout
the mass range. The molecular ion at mass -1-12 and the very
scarce ion ;>t mass 36~> were selected for the same purpose.
Abundant ions at masses (!!). 77, 110, and 2f».r> were not used
because the selected ions adequately measure system sensi-
tivity. Ma»s (?S» wns specifically excluded from the criteria
because its abundance frequently depends on background
conditions thai result from tin? use of J'KK, PFTHA, etc..
for mnsx/charge scale calibrations.
In spectra (Tables 11 and 111) thnt had muss 198 ns tin;
bu«c peak, .seven of the nine molecular ion measurements
Table IV. Reference Compound Key Ions und
ion Abundance Criteria
lou itiundjare criteria
30-60% of m:iss 198
Less than 2'.V. of mass 09
Less than 2% of mass C9
40-60% of mass 198
Less thaji !'£ of mass 198
Base peak, 100'',', relative abundance
5-9% of mass 198
10-30% of mass 198
l%of mass 198
Less than mass 443
Greater than 40% of mass 198
17-23% of mass 442
51
08
70
127
197
198
199
275
365
441
442
443
were greater than 40% relative abundance. Therefore, this
was selected as a reasonable lower- limit for the molecular
ion abundance. No upper 'iir.it was set. All r.ine spectra
showed an ion of 1-3% at mass 365 and a system with ade-
quate high mass sensitivity should detect at least a 1% ion
at this mass. The average abundance for mass 275 in the
nine measurements was 22% with a standard deviation (a)
of 5%. This was rather low dispersion for a set of relative
abundance measurements and suggests that the abundance
at mass 275 is closely related to the arbitrarily constant ion
abundance at mass 198. This is consistent with the compo-
sition assignments discussed previously. However, a toler-
ance at mass 275 of ±5% was considered too small for rou-
tine GC/MS applications. A criterion at mass 275 of 20 ±
10% was selected by rounding the average relative abun-
dance to the nearest ten percent and allowing a deviation of
2. Rounding
off and using a 2
-------
In the mid nntl IMW mass rondos, similar resold! ion checks
were developed i:\inj; ions ronl.'iiijinj; a single i:'C ion at
massed 100 ami 70. In each insl.-mc-u, the ions an very likely
fissioned the correct composition nnd the thmrrticnl per-
centages mny l»c compared with ihc experimental. At mass
190, nine mcaMi.-c:ncnis gave nn average of 8.(X\, nnd a —
2.3. The criterion suggested is 7 ± 2% which compares with
the tKeorcticnl value of 6.6%. At mass 70, ihr iheoretical
value is 1.1%, hut most of the nine measurements nave near
zero values for this ion. Perhaps this wns caused l)y very
slight changes in the base-line (threshold).adjustments. It
is very difficult to make accurate and precise measure-
ments of relatively non-abundant ions when observing very
small amounts (=s20 ng) in fast (3—4 sec) spectrometer
scans. Therefore, for mass 70, we suggest a nominal criteri-
on of less than 2% of mass 69. This is mainly a check on ex-
cessive broadness or poor peak shape in the low mass re-
gion for those data systems that interpret broadness as ion
abundance.
Because of the probable compositions of the mass 198
(CGF5P+) and mass 69 (CF.-1+) ions, it is unlikely that mass
197 and mass C8 ions would he present. Indeec' repeated
measurements have shown that they are -not present.
Therefore, we suggest that mass 197 should be less than 1%
of the base peak, and mass 68 less than 2% of mass 69. Both
criteria are checks on excessive broadness and skew as dis-
cussed above.
CONCLUSION
The set of relative abundance ranges proposed for
DFTPP has been very useful in evaluating the performance
of n number of (JC/MS systems. Those ranees nrc the basis
for the proposed standard ion abundance calibration and
provide a reasonable basis for comparing the output from
the wide variety of systems in use.
ACKNO WLKDC MENT
We express our sincere appreciation to the individuals
and laboratories that participated in the interl.iborntory
study. These included J. Peterson, Fish and Wildlife Ser-
vice; E. M. Chait, K. I. DuPont de Nemours &. Company; J.
C. Cook, University of Illinois; J. B. Knight, Finnifjan Cor-
poration; and F. Biros, J. Blazevich, H. Boyle, M. Carter, P.
Clifford, F. Farrell, G. Muth, H. Rodriguez, D. C. Shew,
and A. Wilson of the Environmental Protection Agency.
LITERATURE CITED
(1) S. S. Qua, R. C. Edmondson. and H. Gilman. J. Organome'-ai. Oem.. 24,
703(1970).
(2) J. VV. EicTielberger. L. E. Harris, ana W L. Budde. to b« y^: >h»3: pre-
sented at th« 22nJ Annual Conference en Mass Spectrome:r, and Aflied
Topics, Philadelphia, PA. May 19-24. 1374
(3) American Society for Testing and Materials Committee 0-2. 21st Annual
Conference on Mass Spectrometry and Al:,ed Topics. San Francisco.
CA, May 20-25. 1973.
(4) B. S. Middleditch. Anal. Chem.. 41. 2092 (1969).
(5) R. H. Wallick. G. L-. Pecle, and J. B. Hyres. Anal. Chem.. 41.382 (1969).
(6) O. M. Schoengold and W. H. Stewart. Anal. Chem.. 44, 834 (1S72).
RECEIVED for review October 29, 1974. Accepted January
24,1975.
-------
CE NO. 3
MEASURING VOLATILE ORGANIC MATTER IN WATER BY AQUEOUS-INJECTION
GAS CHRCMATOGRAPHY
-------
Designation: D 2908 - 74
Standard Recommended Practice for
MEASURING VOLATILE ORGANIC MATTER IN
WATER BY AQUEOUS-INJECTION GAS
CHROMATOGRAPHY1
This Standard it issutd under the Hied designation D 290S: the number immediately following (he designation mdicatesihc
tear of original adoption or. in the case of revision, the >ear of last revision. A number in parentheses indicates the vear of
last reapproval.
1. Scope
1.1 This recommendeo practice covers the
general considerations for the qualitative and
quantitative determination of volatile organic
constituents in water by gas-liquid chromatog-
1.2 Direct aqueous injection of samples is-
feasible at organic concentrations greater than
1 me/liter. The applicability of the method
can be extended to waters of lesser concentra-
tions by evaporative techniques, freeze-out.
solvent extraction, or carbon adsorption.4
2. Significance
2.1 The major organic constituents in in-
dustrial waste water need to be identified for
support of effective in-plant or pollution con-
trol programs. Currently the most practical
means for tentatively identifying and measur-
ing a range of volatile organic compounds is
gas-liquid chromatography. Positive identifi-
cation requires supplemental testing (for
example, multiple columns, speciality detec-
tors. spectroscopv, or a combination of these
techniques).
3. Summary of Method
3.1 This recommended practice defines the
applicability of various columns and condi-
tions for the separation of paturally occurring
or synthetic organic:, or both, in an aqueous
medium for subsequent detection with a flame
ionization detector. After vaporization, the
aqueous sample is carried through the column
by an inert earner gas. The sample compo-
nents are partitioned between the carrier gas
and a stationary liquid phase on an inert solid
support. The column effluent is burned in an
air - hydrogen flame. The ions released from
combustion of the organic components induce
an increase in standing current which is meas-
ured. Although this method is written for hy-
drogen flame detection, the basic technology
is applicable to other detectors if water does
not interfere.
3.2 The elution times are characteristic of
th'e various organic components present in the
sample, while the peak areas are proportional
to the quantities of the components. A discus-
sion of gas chromatography is presented in
ASTM Recommended" Practice E 260. Gen-
eral Gas Chromatography Procedures.6
4. Definitions
4.1 The following terms in this recom-
mended practice are defined in accordance
with ASTM Definitions D 1129, Terms Relat-
ing to Water*:
4.1.1 "ghosting" or memory peaks—an
interference, showing as a peak, which ap-
" This recommended practice is ur»dcr the jurisdiction of
ASTM Committee f>l9on Water.
Current edition approved June 27. 1974. Published July
1974. Originall* published as D :908 - 70 T. Last preview
edition D~:90S'- 70 T.
B> publication of this standard no position is taken »'ih
rc^pe.n to the xjlidit) of an\ patent rights m connection
!T)ere»ith. and the ^mencan Society for Testing and Mate-
rials doe> not undertake to insure anyone utilizing the it-">-
djrd jyjin^i liability for infringement of any Letters Patent
nur j—ume jnv >uch liability.
• The boldface numbers in parentheses refer to the refer-
ences appended to inn recommended onnicc.
Refer aNo (o ASTM Meihod D 24&0, Tesi for Phenol*
m U jter b> Gji-Liyjid Chromotography. which appears'»
this publication
' Kor information on i«o of ihcte concentration teen*
niuues. refer 10 ASTM Meihod D 2778. Solvent Extract**
01' Organic Miner from Water and ASTM Recommended
Practice D XIJ. Removal of Organic Matter from Wiier
K V.l.'jlcJ ( jrrs.n Adsorption, both of »hich appear i»
the lt-4 Annual Bouk of ASTM Standards. Pan 31.
• 1V'4 Annual Hoot of ASIH Standard!. Part 42.
' 1974 Annual Soak of ASTU Slaadanis. Part 31.
480
-------
D 2908
pears at the same elution time as the organic
component of previous analysis.
4.1.2 internal standard—a compound of
known behavior added to a sample to facili-
tate the analyses.
4.1.3 noise—an extraneous electronic sig-
nal which affects baseline stability.
4.1.4 retention time—the lime that elapses
from the introduction of the sample until the
peak maximum is reached.
4.1.5 relatice retention ratio—the retention
time of the unknown component divided by
the retention time of the internal standard.
4.2 For definitions of oiher terms used in
these methods, refer to ASTM Recommended
Practice E 355, Gas Chromatography Terms
and Relationships/
5. Interference
5.1 Paniculate Matter— Paniculate or sus-
pended matter should be removed by centrifu-
gation or membrane filtration if components
of interest are not altered. This pretreatment
will prevent both plugging of syringes and
formation of condensation nuclei. Acidifica-
tion will often facilitate the dissolving of par-
ticulate matter, but the operator must deter-
mine that pH adjustment does not alter the
components to be determined.
5J Identical Retention Times—With any
given column and operating conditions one or
more components may elute at identical re-
tention times. Thus a chromatographic peak
is only presumptive evidence of a single com-
ponent. Confirmation requires analyses with
other columns with varying physical and
chemical properties or spectromethc confir-
mation of the isolated peak or both.
5.3 Acidification—Detection of certain
groups of components will be enhanced if the
sample is made neutral or slightly acidic. This
may minimize the formation of nonvolatile
salts in cases such as the analysis of volatile
organic acids and bases and certain chloro-
phenols.
5.4 Ghosting—Ghosting is evidenced by an
interference peak that occurs at the same time
ts that for a component from a previous anal-
ysis but usually with less intensity. Ghosting
occurs because of organic holdup in the injec-
tion port. Repeated water washing with 5-»l
injections between sample runs will usually
eliminate ghosting problems. The baseline is
checked at maximum sensitivity to assure that
the interference has been eliminated. In addi-
tion to water inieciions, increasing the injec-
tion port temperature for a period of time'will
often facilitate the elimination of ghosting
problems.
5.4.1 Delated Elution—Highly polar or
high boiling components may unpredictably
elute several chrornatograms later and there-
fore act as an interference. This is particularly
true with complex industrial waste samples. A
combination of repeated water injections and
elevated column temperature will eliminate
this problem. Back flush valves should be used
if this problem is encountered often. Carrier
gas wetted by steam can be used to reduce
component holdup in some cases: however.
column life may be seriously shortened. Pass-
ing the carrier gas through a pre-column con-
taining copper sulfate (CuSO4 • 5H;0) for
wetting may have a lesser effect on substrate
stripping (1).
6, Apparatus
6.1 Gas System:
6.1.1 Gas Regulators— High quality pres-
sure regulators should be used to ensure a
steady flow of gas to the instrument. If tem-
perature programming is used, differential
flow controllers should be installed in the car-
rier gas line to prevent a decrease in flow as
the pressure drop across the column increases
due to the increasing temperature. An un-
steady flow will create an unstable baseline.
6.1.2 Gas Transport Tubing—New rubing
should be washed with a detergent solution.
rinsed with cold water, and solvent rinsed to
remove residual organic preservatives or lu-
bricants. Ether is an effective solvent. The
tubing is then dried by flushing with nitrogen.
6.1.3 Gas Leaks—The gas system should
be pressure checked daily for leaks. To check
for leaks, shut off the detector and pressurize
the gas system to approximately 103 kPa (15
psi) above the normal operating pressure. Then
shut off the tank valve and observe the level of
the pressure eauge. If the preset pressure holds
for 10 min, the system can be considered leak-
free. If the pressure drops, a leak is indicated
and should be located and eliminated before
proceeding further. A soap solution may be
481
-------
D 2908
used for determining the source of leaks, but
care must be exercised to avoid getting the
solution inside the tubing or instrument since
it will cause a long lasting, serious source of
interference. Leaks may also occur between
the instrument gas inlet valve and flame tip.
This may be checked by removing the flame
tip, replacing it with a closed fitting and re-
checking for pressure stability as previously
noted.
6.1.4 Gas Flow— The gas flow can be deter-
mined with a bubble flow meter. A micro-ro-
tameter in the gas inlet line is also helpful. It
should be recalibrated after each readjustment
of the gas operating pressure.
6.2 Injection Port—The injection port
usually is insulated from the chromatographic
oven and equipped with a separate heater that
will maintain a constant temperature. The
temperature of the injection port should be
adjusted to approximately 50 C above the
highest boiling sample component. This will
help minimize the elution time, as well as
reduce peak tailing. Should thermal decompo-
sition of components be a problem, the injec-
tion port temperature should be reduced ap-
propriately. Cleanliness of the injection port
in some cases can be maintained at a tolerable
level by periodically raising the temperature
25 C above the normal operating level. Use of
disposable glass inserts or periodic cleaning
with chromic acid can be practiced with some
designs. When using samples larger than 5 ul.
blowback into the carrier gas supply should be
prevented through use of a preheated capillary
or other special design. When using 3.175-mm
(0.125-in.) columns, samples larger than 5 n\
may extinguish the flame depending on col-
umn length, carrier gas flow, and injection
temperature.
6.2.1 Septum—Organics eluting from the
septum in the injection port have been found
to be a source of an unsteady baseline when
operating at high sensitivity. Septa should be
preconditioned. Insertion ot" a new septum in
the injection port at the end of the day for
healing overnight will usually eliminate these
residuals. A separate oven operating at a tem-
perature similar to that of the injection port
can also be used to process the septa. The
septa should be changed at least once a day to
minimize gas leaks and sample blowback.
Septa with TFE-fluorocarbon backings mini-
mize organic bleeding and can be used safely
for longer periods.
6.2.2 On-Column Injection—While injec-
tion into the heated chamber for flash vapori-
zation is the most common injection set-up.
some analyses (for example, organic acids)
are better performed with on-column injection
to reduce ghosting and peak tailing and to
prevent decomposition of thermally degrada-
ble compounds. This capability should be built
into the injection system. When using on-col-
umn injection a shorter column life may occur
due to solid build up in the injection end of
the column.
6.3 Column Ocen—The column ovens
usually are insulated separately from the
injection port and the detector. The oven
should be equipped with a proportional heal
and a squirrel-cage blower to assure maxi-
mum temperature reproducibility and uni-
formity throughout the oven. Reproducibility
of oven temperature should be within 0.5 C.
6.3.1 Temperature Programming—Tem-
perature programming is desirable when the
analysis involves the resolution of organics
with widely varying boiling points. The col-
umn oven should be equipped with tempera-
ture programming between 50 and 350 C with
selectability of several programming rates
between 1 and 60 deg/min provided. The ac-
tual column temperature will lag somewhat
behind the oven temperature at the faster
programming rates. Baseline drift will often
occur because of increased higher tempera-
tures experienced during temperature pro-
gramming. This depends on the stability of
the substrate and operating temperature
range. Temperatures that approach the maxi-
mum limit of the liquid phase limit the oper-
ating range. Utilization of dual matching col-
umns and a differential electrometer can min-
imize the effect of drift: however, the drift is
reproducible and does not interfere with the
analysis in most cases.
6.4 Detector—The combination of high
sensitivity and a wide linear range makes the
flame ionization detector (FID) the usual
choice in trace aqueous analysis. The flame
ionization detector is relatively insensitive to*
water vapor and to moderate temperature
changes if other operating parameters remain
unchanged. If temperature programming is
used, the detector should be isolated from the
482
-------
flb
D 2908
oven and heated separately to ensure uniform
detector temperature. The detector tempera-
ture should be set near the upper limit of the
programmed temperature to prevent conden-
sation. The detector should also be shielded
from air currents which could affect the burn-
ing characteristics of the flame. Sporadic
spiking in the baseline indicates detector con-
tamination: cleaning, preferably with diluted
hydrochloric acid (HC1. 5 4- 95), and an ultra-
sonic wash with water is necessary. Chromic
acid also can be used if extreme care is taken
to keep exposure times short and if followed
by thorough rinsing. Baseline noise may also
be caused by dirty or corroded electrical con-
tacts at switches due to high impedance feed-
back.
6.5 Recorder—A 1-mV, l-s, full-scale re-
sponse, strip-chart recorder, is recommended
to obtain a permanent chromatogram. Chart
speeds should be adjustable between 15 and
90 in./h.
6.6 Power Supply—A. 105 to 125-V, a-c
source of 60-Hz frequency suppling 20-A serv-
ice is required as a main power supply for
most gas chromatographic systems. If voltage
fluctuations affect baseline stability, a voltage
regulating transformer may be required in
addition to the one incorporated within the
chromatographic instrument.
7. Retgents and Materials
7.1 Purity of Reagents—Reagent grade
chemicals shall be used in all instances for gas
purification, sample stabilization, and other
applications. Unless otherwise indicated, it is
intended that all reagents shall conform to the
specifications of the Committee on Analuical
Reagents of the American Chemical Society.
where such specifications are available.' Other
grades may be used, provided it is first ascer-
tained that the reagent is of sufficiently high
purity to permit its use without lessening the
accuracy of the determination.
7.1.2 All chemicals used for internal stand-
ards shall be of highest known punty.
7.2 Purity of Water—Unless otherwise in-
dicated, references to water shall be under-
stood to mean reagent water conforming to
Type I of ASTM Specifications D 1193. for
Reagent Water.'
7.3 Carrier Gas System—Only gases of the
highest purity obtainable should be used in a
chromatographic system designated for trace-
organic monitoring in water. The common
carrier gases used with a flame ionization de-
tector (FID) are helium and nitrogen. Trace
contaminants in even the highest purity gases
can often affect baseline stability and intro-
duce noise. Absorption columns of molecular
sieves (14 by 30-mesh) and anhydrous calcium
sulfatc (CaSO4, 8 mesh) in series between the
gas supply tank and the instrument will mini-
mize the effect of trace impurities. These pre-
conditioning columns, to remain effective,
must be cleaned by back flushing them with a
clean gas (nitrogen, helium) at approximately
200 C. or they must be replaced at regular
intervals. Use of catalytic purifiers is also
effective (4).
7.4 Column:
7.4.1 Column Tubing— For most organic
analyses in aqueous systems, stainless steel is
the most desirable column tubing material.
However, when analyzing organics that are
reactive with stainless steel, glass tubing
should be used. With a flame ionization detec-
tor, maximum resolution with packed columns
is achieved with long, small-diameter (3.175-
mm (0.125-in.) and smaller) tubing. New tub-
ing should be washed as described in 6.1.2.
7.4.2 Solid Support—Maximum column
efficiency is obtained with an inert, smail.
uniform-size support. The lower limit of parti-
cle size will be determined by the allowable
pressure drop across a column of given diame-
ter and length. Elimination of fines will re-
duce the pressure drop and allow the use of
smaller panicles: the commonly used size is
80/100 mesh. Supports, which are not inert,
may cause varying degrees of peak tailing.
Few supports can be classified as totally inert:
however, techniques are available to assist in
the deactivation of the support. Chromosorb
"W",* the least active type of diatomaceous-
earth support, can be further deactivated by
acid or base washing. A combination of acid
washing and silanization (for example, dimcth-
yldichlorosilanc (DMCS), hexamethyldisilane)
'"Reagent Chemicals, American Chemical Society
Specifications." Am. Chemical Soc.. Washington. D C.
For suggestions on the letting of reagents not listed by the
American Chemical Society, see "Rcaeent Chemical! and
Standard!." h> Joseph Rosin. D. Van Nostrand Co., Inc..
New York. N.Y.. and the "United States Pharmacopeia."
* This material, while proprietary in nature, a distinctly
superior to others which have been tried and is available
from essentially all vendors of chromalogrxphic supplies.
483
-------
O 2908
treatment may reduce the surface activity still
further. However, silanization can decrease
column life. DMCS treatment is particularly
useful when low bquid leads are used. Treat-
ment with specific chemicals that approximate
the properties of the sample being analyzed
has also proven successful. For example, ter-
cphthalic acid treatment of Carbowax 20M*
reduces organic acid and phenolic tailing. Use
of fluorocarbon supports can significantly re-
duce tailing. For low boiling materials, porous
polymer beads formed by the polymerization
of monomers such as styrene wiih divinyi ben-
zene as a crosslinker are finding more applica-
tion in trace analysis. Since there is no liquid
phase, there is minimal column bleed during
temperature programming. In addition, elimi-
nation of the conventional solid support re-
moves the adsorptive sites which normallx
cause tailing. Caution must also be taken not
to exceed the recommended maximum tem-
perature limit of the fluorcarbon supports or
of the porous polymer beads being used.
7.4.3 Liquid Phases—Maximum resolution
and minimum baseline noise and drift are
achieved with a relatively lightly loaded col-
umn, (less than 5 percent) containing a stable
substrate of low volatility. However, analysis
of aqueous samples with light column loading
produces shorter column life and a greater
tendency for a shift in retention times and
delayed elution as the column ages. Acceler-
ated aging will occur if the maximum temper-
ature limit of the liquid phase is exceeded or
approached repeatedly. Substrates should be
selected to permit operation at a temperature
below the maximum allowable if at all possi-
ble. Selection of liquid phases should be based
on the properties of the sample to be ana-
lyzed. In general, polar substrates will resolve
polar compounds b> order of relative volatility
and polarity. Polar substrates will resolve
nonpolar comoounds by structural t>pe. Non-
polar substrate:- »iil separate nonpolar com-
pounds b\ volatility and polar compounds by
structural type, r-or examples of applicable
liquid phases for a particular application, con-
sult published methods for specific organic
classes.
7.4.4 Column Conditioning All new col-
umns should be pre-conditioned to drive off
the rcMdual contaminants which would foul
the detector and cause severe baseline noise.
New columns can be conditioned by attaching
one end to the inlet port of the oven and al-
lowing 20 to 30 ml/min of carrier gas to pass
through the column either at 30 C above the
expected maximum operating temperature or
at the maximum temperature limit of the liq-
uid phase, whichever is lower. The effluent
end of the column should be vented. The col-
umn should not be attached to the detector
during conditioning since cluting organics
may foul the detector. Occasional 5-drogen and air of
the highest initial purity which have been fur-
ther purified as described in 7.3. are fed to the
detector. Hydrogen can also be used which is
produced from the electrolytic decomposition
of water.
7.6 Glassware—All glassware that will
come into direct contact with the sample
should be heated in an oven to 300 C
(overnight if possible) as a final cleanup step.
This will serve to remove any source of or-
ganic contamination from prior work.
8. Samples and Sampling Procedure
3.1 Sample Collection—Collect all samples
in accordance with the applicable method of
the American Society for Testing and Materi-
als as follows:
D 510—Sampling Water,*
D 1192—Equipment for Sampling Water
and Steam.'and
D 1496—Sampling Homogeneous Waste
Water.*
Additionally sample containers and sample
MZC and storage shall be as specified in 8.2 to
8.4.
8.2 Sample Containers—CATC should be
taken to collect a representative sample in a
clean, completely full glass bottle. The screw
cap should be lined with aluminum foil or
TFE-fluorocarbon to reduce the sorption of
484
-------
D2908
insoluble organics.
8.3 Sample Sire—The sample size must be
small lo prevent overloading of the 3.175-mm
(0.125-in.) columns generally used. For most
aqueous analyses, a sample size of 2 to 5 u\ is
generally optimum. If the components of in-
terest are of relatively high concentration, a I-
>t\ sample is to be used. At low concentra-
tions, a sample approaching 10 «il can be
used to increase the detectable limit although
the measurement accuracy is slightly decreased
since a 10-pl syringe is used. For the best ac-
curacy, select a syringe with a capacity 50
percent greater then the size of the sample to
be injected.
8.4 Sample Storage—Storage time of sam-
ples should be kept to a minimum. If storage
cannot be avoided, the bacterial action should
be minimized by refrigeration, by pH adjust-
ment to about 2.0 (if organics are not acid
degradable). or by the addition of 1 ml of sat-
urated mercuric chloride (HgCh) solution to
each liter of sample. Selection of a preserva-
tion procedure is dependent on the analysis
being made.
9*. Preparation of Chromatograph
9.1 Column—Select the appropriate col-
umn and install in the chromatographic oven.
If the column is new. it should be precondi-
tioned according to the directions in 7.4.4.
The column should then be attached to the
detector and the system checked for leaks
according to 6.1.3. The column temperature
requirements should be set according to the
requirements outlined in the specific method
being used.
9.2 Gases—With a flame ionization detec-
tor the gases require adjustment in the ratio
of about 1 part carrier gas to 1 part hydrogen
to 10 parts air. A typical flow for the carrier
gas when using 3.175-mm(0.125-in.) tubing is
25 ml/tnin. Refer to the specific method
being used for flow requirements.
9.3 Electrometer and Recorder—Adjust the
electrometer and recorder as specified on the
instrument instructions so that the pen is ze-
roed and the attenuation steps are linear.
Based on the organic content of the sample to
be analyzed, adjust the electrometer attenua-
tion to give as near mid-scale deflections of
the recorder pen as is practical.
9.4 Baseline Stability- Before proceeding
with the analysis, check the stability of the
recorder baseline with the pen at zero and the
attenuation at the level to he used for the
analysis. If sporadic peaks occur, further col-
umn conditioning may be necessary. The re-
corder, electrometer, flow controllers, and
flame detectors should also be checked as a
possible source of the sporadic peaks.
9.5 Column Storage— When columns are
not in use. their ends should be capped. The
need for reconditioning prior to their reuse at
a later time will be indicated by making cali-
bration runs with a known concentration of
standards. Reconditioning is generally mini-
mal if the column was adequately purged
prior to storage.
10. Calibration and Standardization
10.1 Qualitative:
10.1.1 The basic method of tentative com-
pound identification is by matching the reten-
tion times of known standards suspected to be
present with retention times of unknown com-
pounds under identical operating conditions.
The absolute retention time is measured in
minutes from the time of injection to the peak
maximum. Since retention time may van
significantly with concentration of the particu-
lar organic compounds, identification is done
more positively by spiking the sample with
the suspected constituent and noting an in-
crease in peak height. In some instances more
than one compound may elute at the same
time and therefore have identical retention
times. This condition can often be recognized
by a poorly shaped peak (that is. double apex
or shoulder). When this occurs, additional
column(s) with different physical and chemi-
cal properties will be required to separate the
combined peaks. An alternative, which is fre-
quently preferable, is to trap the peaks and
identify them spectromctrically (see 11.7V
10.1.2 Relative retention times are devel-
oped by the insertion of a common noninter-
fering organic into each standard as well as
into the unknown. The absolute retention time
of the common organic is then divided into
the absolute retention time of each organic
being analyzed. Utilization of relative reten-
tion times improves qualitative accuracy by
balancing out numerous chromatographic var-
iations from run to run. for example, slight
variations in column temperature, program-
485
-------
D 2908
ming rate, carrier gas Row, and sample size as
well as column aging.
10.1.3 Based on the type and concentration
of compounds expected in the sample to be
tested, prepare similar standards in reagent
water.
10.1.4 At least three relative retention
times with a single column should be deter-
mined for each organic standard and the aver-
age used for qualitative analysis of the un-
known sample. Relative retention times
should be verified periodically.
10.1.5 One and two-column identifications
are not usually sufficient for positive identifi-
cation. A third column or spectrometric analy-
sis of the trapped peak will be required for an
unequivocal identification.
10.2 Quantitative:
10.2.1 The quantitative measurement of
each component is determined from the area
under the individual chromatographic peaks.
Peak areas can be determined more effi-
ciently by mechanical or electronic integra-
tors. If the peaks are symmetrical and sharp
with minimum tailing, peak height can be
used for estimating quantitative response for
expediency in routine monitoring type analy-
sis. The height is measured from the peak
maximum to the baseline. If the peak occurs
in an area of baseline drift, approximate the
actual base of the peak for measuring pur-
poses. Measuring the peak width at one half
the peak height and multiplying it by the peak
height will approximate the peak area. The
error increases as the peak width becomes
smaller or as peak tailing increases.
10.2.2 Insertion of an internal standard is
useful for quantitative analysis. When re-
sponse is calculated relative to an internal
standard, compensation is provided for the
inadvertent changes in chromatographic con-
ditions. Selection of the internal standard
should be based on its separation from other
peaks, stability, and if possible on mid-chro-
matogram elution and structural similarity to
the components being analyzed. The internal
standard should be applied at approximately
the expected average concentration of the
organic constituents. When temperature pro-
gramming is used, two internal standards inny
be needed, one for low-boiling and one for
high-boiling constituents.
10.2.3 Mass response ratios are determined
by the injection of standards containing the
same concentration of both the internal stand-
ard and the individual components suspected
to be in the samples to be tested. For accurate
quantitative work triplicate injections should
be made on a conditioned column with the
average being used for further calculations.
All chemicals used should be of the highest
known purity, so that the appropriate correc-
tion may be made when calculating the final
response factors. Response factors should be
rechecked periodically.
10.2.4 The linearity of the response factors
should be verified by varying the concentra-
tion of the individual components over the
concentration range of interest while holding
the internal standard concentration constant.
These ratios when plotted against concentra-
tion should yield a straight line that passes
through zero. Chromatographic operating
conditions should always be recorded on the
graph. Attenuation should preferably be ad-
justed to keep the peaks at approximately 50
percent of full scale, if possible. The final
peak areas or heights are adjusted according
to the electrometer attentuation setting used
for calibration.
11. Sample Testing Procedure
11.1 Injection Practice—Use a firm, rela-
tively fast injection technique so that the
sample can be injected either into the middle
of the injection port for flash vaporization, or
approximately 2 in. down the column for on-
column injection in a slug-like condition. Slow
injections may cause poor resolution and
spreading. Use the same rhythm each time.
Wash the syringe several times between injec-
tions with acetone, then rinse with water, and
air dry by attaching to a vacuum line. Flush
the syringe at least two times with the sample
to be analyzed. Remove the bubbles by pump-
ing the syringe plunger followed by a slow
drawup of the sample. When injecting large
samples at high inlet pressure (for example.
50 psi). hold the plunger so as to prevent its
blowout caused by the pressure buildup in the
injection port: special syringes are needed for
high-pressure work.
11.1.1 Sample Injection—Use direct
aqueous injection whenever possible to pre-
486
-------
D 2908
vent both the loss of some component-, and the
introduction of extraneous peaks .that may
result from concentration techniques How-
ever, when analyses are in the part per billion
range, concentration techniques will be re-
quired. Carbon adsorption, gas stripping, sol-
vent extraction, and freezeout have been
shown to increase component concentration to
detectable levels«IJS.6\.
11.2 Establish operating conditions identi-
cal to those used for calibration and standard-
ization. If changes are required because of
sample peculiarities, repeat calibration and
standardization using the new conditions. If
an internal standard is used, minor changes in
operating conditions are tolerable.
11.3 Inject sample prior to insertion of in-
ternal standard to assist in either the selection
of the internal standard, or to assure that the
internal standard selection is well resolved
from component peaks in the sample. An open
position in the chromatogram is selected for
this purpose.
11.4 Add the internal standard(s) into the
sample at a concentration approximating the
components to be analyzed and repeat the
analysis.
11.5 Refer to the specific method for sug-
gested sample size: 3 to 5 ul are often used.
11.6 Determine the absolute retention
times of the individual components in the
sample. Calculate relative retention times
using the retention time of the internal stand-
ard in the denominator. Refer to the pre-
viously developed listing for relative retention
times of known compounds on specific col-
umns: if absolute retention times are used.
run standards several times during the test
series. Repeat on additional columns as neces-
sary to increase qualitative accuracy.
11.7 Trap individual peaks for confirma-
tory analysis. Mass spectromctric analysis of
trapped components is often most informa-
tive: however, infrared spectroeraphic analy-
sis, thin-layer chromatography. and microcou-
lometry or other speciali/ed detectors (for
example, flame photometric detector, modi-
fied flame halogen detector) are iil Gas-
Liquid Chromatoeraphv." International Journal
of Air and Water Pollution. UPWA, Vol 26.
1966. pp. 591 to 602.
|J) Baker. R. A.. "Volatile Faiu Acids in Aqueous
Solution by Gas-Liquid Chromatoeraphx."
Journal of Gas Chnjniaii>tmph\. JGCRA. Vol
4. 1966. pp. 41$ to4|<)
|4> Baker. R. A. and Malo. B. A.. "Phenolic* b%
Aqueous-Injection Gas Chromatographv."
Journal of Environmental Science and Techftol-
of\\ Vol I. 1967. pp. 997 to 1007.
(5» Baker. R. A.. "Trace Organic Contaminant
Concentration by Freezing-l; Low Inorganic
Aqueous Solutions." Journal of the lmt*na-
tional Association on Water Pollution Research
Vol I. 1967. pp. 61 to 77.
t6t Baker. R. A.. "Trace Organic Contaminant
Concentration by Freezing-11: Inorganic
Aqueous Solutions." Journal of the Interna-
tional Association on Water Pollution Re-
search, Voll. 1967.pp.97to MJ.
487
-------
NO. 4
NTTRILES IN AQUEOUS SOLUTION BY GAS-LIQUID CHROMATOGRAPHT
-------
Designation: O 3371 - 74 T
Tentative Method of Test for
NITRILES IN AQUEOUS SOLUTION BY GAS-LIQUID
CHROMATOGRAPHY1
Thi* Tentative Method has been approved by the sponsoring committee and accepted by the Society in accordance with
established procedures, for use pending adoption as standard. Suggestions Tor revisions should be addressed to the Society, at
1916 Race St.. Philadelphia. Pa. 19103.
1. Scope
1.1 This method covers nitriles that can be
separated and detected quantitatively at a limit
of approximately 1 mg/litre by aqueous injec-
tion on a selected gas-liquid chromatographic
column.
1.2 This method utilizes the procedures and
precautions as described in Recommended
Practice D 2908.
2. Applicable Documents
2.1 ASTM Standards:
D 2908 Recommended Practice for Measur-
ing Volatile Organic Matter in Water by
Aqueous Injection Gas Chromatography*
3. Significance
3.1 Nitriles at concentrations of a few milli-
grams per litre are potentially toxic to aquatic
life. Nitriles in waste water discharges should
be detected and controlled.
3.2 Gas-liquid chromaiography (GLC) can
detect and determine mixtures of nitriies at
levels where wet chemical procedures are not
applicable.
4. Special Comments
4.1 It is recommended that samples that
cannot be analyzed immediately, be quick
frozen for preservation. Samples should be
aeutralized to pH 7 at the time of collection to
minimize hydrolysis of the nitrite groups.
4.2 Samples of nitriles to be employed as
standards should be considered to be unstable.
Storage in a freezer is recommended.
4.3 It is not always practical to translate
operating conditions directly from one GLC
instrument to another. An operator should
optimize his instrument to a particular proce-
dure, for example, injection and detection tem-
perature, flow rates, etc.
5. Typical Chromatograms
5.1 The following instrument parameters
were used to obtain the typical chromatograms
(See Figs. 1 and 2).
5.1.1 Column—vs in. outside diameter
stainless steel, 8 ft long packed with a porous
styrene divinylbenzene polymer.
NOTE—"Chromosorb" 101. 50/60 mesh, was used
for tbe typical chromatograms.
5.1.2 Detector, flame ionization.
5.1.3 Temperatures:
Injection port 240*C
Detector 240*C
Oven, isothermal 130*C
Oven, programmed at llO'Ciomax
10'C/min of200*C
5.1.4 Carrier Gas, helium at 25 ml/min.
5.1.5 Sample Size:
isothermal 5 *il
programmed 3 u\
5.1.6 Recorder, 3/i in./min chart speed and 1
mV full-scale response.
5.2 Kovats Index Values:*
Compounds
Acetonitrile
Acrylonitnle
Relative
Retention
1.00
Kovats
Index
470
512
1 This method ii under the jurisdiction of ASTM Commit-
tee 0-19 on Water and is the direct responsibility of Subcom-
mittee D 19.05 on Inorganic Constituents in Water.
Current edition approved Nov. 4, 1974. Published Feb-
ruary 1975.
'Ann-jal Soot of ASTM Standards. Part 31.
' Gas Chromatographic Data Compilation, ASTM
AMD 23A. Am. Sot Testing Mats.. 1967.
534
-------
D3371
Propnoortnie
Iwjvtkronitnk
Vileronnnle
Hcuaenimlc
Beiuonitnle
1.67
2.1\
I.JO
304
3.JS
4.2i
5.42
570
635*
678
740'
783
90S*
990
6. Precision
6.1 The precision of this method within the
range from 10 to 60 mg/litre of standards in
distilled water may be expressed as follows:
Compound
Acetonilnle
Propiominle
Meihoxy Acetonitnlc
Butyronitrile
where:
Sr
mg/ litre
5r
Sr . O.OI5(mg/lure) -t- 0.9
5r - 0.088 (mg/litre) - 0.6
Sr - 0.097(mg/hlre) -r O.I
ST - O.IO(mg/!ilre) -0.4
overall precision, and
concentration of the specific com-
pound
' Kov*u index vilues estimated from relative retentioa
data became standard compound was not readily available.
Tine IH
J/*-(rleh a*r «inut«)
Column Packing - Oiromosorb IOI, 50/60 mesi
Corner Ga»- Helium at IS ml/mm
Temperature - Isothermal operation of the column at I30*C
Sample Size - 5 microlilen containing 10 mg/l of each niinie
FIG.
535
-------
03371
3
a
nrraricii TI-C
(Clwrt
l-iief> a..- iirv.t.1
Column Packing • Chrotnosorta 101. 50/60 mob
Carrier Gas- Helium at Z3 mi/min
Temperature- Programmed operation at 10'C/min from 110'C to a maximum of 200*C
Sample Size-} microliten containing l.iOO mg/lof each nitnle
FIG. 2— PropoMd Tcaptncart CorooKognpbk ABaJyju of Ninila in Aqumo* Soivdo*
Tkr Amtritm Societv far Testing aid Mattriais lakes no position respecting the validity of any patent ngnu asserted
in eonneeuon vuh any item mentioned in this standard. L'stn of this standard are expreaiy advised thai determination oftht
validity of any such patent nghu. and the risk of infringement of such rights, is entirety their own responsibility.
536
-------
CE NO. 5
DIRECT ANALYSIS OF WATER SAMPLES FOR ORGANIC POLLUTANTS WITH
GAS-CHRCMATOGRAPHY--MASS SPECTROMETRY
-------
Direct Analysis of Water Samples for Organic Pollutants with
Gas Chromatography-Mass Spectrometry
Lawrence E. Harris, William L. Budde,1 and James W. Elchelberger
Envronmeniai Protection Agency. National Environmental Research Center. Methods Development and Quality Assurance ftose+rcft
laboratory. Cftcmnatt, Otuo 45268
A direct aqueous injection gas chromaiography-mass spec-
Irometry (GC/MS) procedure was explored as a supple-
ment to conventional solvent extraction for analysis of the
organic pollutants in water and waslewater samples. Stud-
ies were made of the effects of relatively large pressures of
water vapor on the well established electron impact frag-
mentation patterns, quadrupole GC/MS system perfor-
mance, interactive background subtraction, and detection
limits. It was concluded that direct aqueous analysis is a
valuable supplemental procedure for the detection of vola-
tile compounds that are not found with solvent extraction.
Effective water pollution control requires analytical
methodology that is capable of generating correct identifi-
cations and measurements of the concentration of the or-
ganic pollutant? in water samples. This methodology is
necessary to determine the exact sources of pollution, to set
effluent standards for toxic pollutants, to enforce effluent
guidelines, to evaluate the effectiveness of treatment facili-
ties, and to determine the causes of taste, odor, and fish
kills.
In the past, a very significant amount of research, fre-
quently over several weeks or months, was required to ob-
tain identifications of the trace organics in water samples.
Often this effort resulted in just a few identifications and it
occasionally produced erroneous results. The development
of computerized gas chromatography-mass spectrometry
(GC/MS) revolutionized the field of trace organic analysis
(7, 2). Today many laboratories have the capability to
make more than a dozen unambiguous identifications with
just a few man-hours of effort.
The first sample work->up methods used with GC/MS in
organic water pollutant analysis were minor modifications
of standard solvent extraction procedures which were de-
veloped for pesticide analyses. These procedures together
with GC/MS are very effective in isolating, concentrating.
and identifying extraciable and volatile organic pollutants
at levels as low as 10 parts per trillion ilO ng/1). This great
sensitivity is achieved, in part, by an efficient concentra-
tion of a relatively large volume of organic solvent extract
to a very small volume. Concentrations of trace organics by
a factor of 10" is not uncommon.
Solvent extraction? doe>. however, possess several limita-
tions including the lo^s of very volatile organic pollutant;
(e.C.. chlorinated solventsi by vaporization during the ex-
tract concentration step. Another difficulty is the failure t<>
extract efficiently a varietv of volatile but water soluble or-
ganic pollutants ie s . low molecular weight alcohols and
ketone solvents). A supplemental work-up procedure for
1 To whom correspondence should be addressed.
(1) ft. A. HrtesandK 8«mann. Sdf.tca. 178, 158(1972).
(2) J. A" EicneOcfcer. L. £. warns, and VV L Budde. Antl. Chem.. 46. ?rT
(1574).
the analysis of these compounds is required. Perhaps the
simplest and most direct approach is the analysis of the un-
altered water samples by GC/MS.
The gas chromatography of unaltered water samples is
feasible and has been known for some time (3-7). It has
been practiced only sparingly, however, because conven-
tional GC detectors (e.g.. the flame ionization detector) do
not produce sufficient information to unequivocally distin-
guish among the enormous variety of different organic
compounds that could be present in a water sample. Com-
puterized GC/MS overcomes this difficulty since the mass
spectrometric data are frequently sufficient for an unam-
biguous characterization of most of the very volatile com-
pounds present.
Routine direct aqueous GC/MS analysis for organic-
water pollutants offers the potentially very significant ad-
ditional benefit of instant analysis. Since no time and
labor consuming pre-analysis treatment is required, a rela-
tively large number of samples may be processed per unit
of time at a relatively low unit cost.
A study was made of the applicability of this technique
to water pollutant identification.
Difficulties that might be anticipated with water as a sol-
vent for GCA1S analysis were studied also. Since the sol-
vent extract concentration step was eliminated, sensitivity
limitations were defined. Traditionally water is considered
highly detrimental in magnetic deflection mass spectrome-
ters. It may facilitate discharges from 2-8 kV accelerating
potentials and cause degradation of Cu-Be electron multi-
plier detectors. However the effects of large quantities (1-
10 pi) of water injected into the GC/MS on the perfor-
mance of the quadrupule mass spectrometer and the sam-
ple enrichment device were unknown. Also unknown was
the effect of large quantities of water vapor on the well es-
tablished electron impact fragmentation patterns of organ-
ic compounds.
The approach used was to analyze representative waste
samples and well-defined mixtures of compounds to ascer-
tain the effect of water on the system and fragmentation
patterns. The levels of detection of a variety of classes of
compounds were determined with several new and old GC
column packing materials.
EXPERIMENTAL
Instrumentation. The water samples were analyzed using di-
rect on-column injection into a Varian Model 1400 gas chrnm.ito-
graph interfaced with a Kinnipan Model 1015 C quadrupole moss
spectrometer system controlled by a Systems Industries 120 data
acquisition and control system (2). The MS wat the only detector
used. The data were displayed as plots on a cathode ray tube dis-
(3) R. C Dress/nan. J. Chromarogr. Sci.. 8. 265 (1970).
(4) F I. Onuska. Water Res.. 7. 835(1973).
(5) M. E. Foi. Snvron. Sci. Technol.; 7, 838 (1973).
(«) J. W. Sugar and R. A. Conway. J. w*ier PotM. Contr. Feet, 40, 1622
(1968).
(7) "Annual Book of ASTM Standards." ASTM. Part 23. Philadelphia, Pa., c
1972. pp 706. fl 19.
1912 • ANALYTICAL CHEMISTRY, VOL. 46. NO. 13, NOVEMBER 1974
-------
Table I. Selected Organic
Compound
Blank
n-Decane
n-Decane
MIBK
MIBK
n -Butyl acetate
n -Butyl acetate
M-Amyl alcohol
M-Amyl alcohol
p-Cresol
^-Cresol
Acetophenone
Acetophenone
2-Phenylethanol
2-Phenylethanol
w-Hexadecane
w-Hexadecane
sec -Butyl alcohol
Acetone
Methyl -n -octanoate
Chloroform
DME
H-Amyl alcohol
M-Amyl alcohol
'Methylene chloride
Methylene chloride
Ethyl acetate
DME
MIBK
Dioxane
Acetophenone
o-Chlorophenol
w-Cresol
Compounds
Puutltv
•9
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
20
5
10
10
5
50
50
50
50
50
50
50
Analyzed by CC/MS
Quantity
•olvcot.
»L
1
1
1 •
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
10
10
1
1
1
1
1
1
1
1
Sohrat
Water
Acetone
Water
Acetone
Water
Acetone
Water
Acetone
Water
Acetone
Water
Acetone
Water
Acetone
Water
Acetone
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Water
Concentration,
019/L
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
10
5
1
1
5
50
50
50
50
50
50
50
cc
column
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
CC column
condition)
a
b
b
b
b
b
b
b
b
c
c
c
c
c
c
c
c
d
d
d
a
a
a
a
a
a
e
c
e
e
e
e
e
« 100" isothermal;' 70* for 2 min. then 6*/min to 120*;' 150* for 2 min; then 6*/min to ISO": •* 100* for 1 min. then 6*/cnin to 150'; • 60* for
3.5 min, then 12' /min to
play unit (Tektronix Model -1010) or a flat-bed plotter (Houston
Instruments Model DP-1). The GC/MS interface utilized an all
glass jet-type enrichment device to deliver the sample directly into
the ion source of the MS. The butch iniet system was all glass with
a constant-leak opening that introduced the sample directly into
the ion source of the MS. The batch inlet was heated to 100° for
the analyses. Other conditions that were held constant throughout
the analyses were: helium carrier gas at a flow rate of about 30
ml/min; temperature of the GC injection port at 190"; the temper-
atures of the interface and transfer line at 210°; detector manifold
temperature at 100": pressure in MS of 10-i Torr; ionizing voltaee
of 70 eV; A filament current of 500 uA: electron multiplier at .1000
volts; mass rar.pe scanned from 'JO-'JOO amu at an integration time
of 10 msec.'smu: and sensitivity at 10~7 A/Volt unless otherwise
specified. The quadrupolc MS operating parameters were adjusted
to give the normal ion abundances for a standard reference Com-
pound (S).
Gas Chromntocraphv Columns. Column ]. .\ 12-foot coiled
stainless steel lo.d. ll.ll.'"> in.) tulxr WHS parked with (>0'SO mesh CAS
Chmwn Q ctiulrd with .v\. CartHiwax ?lk M.
Column 2. A 7-loot coiled class column li.d. 2 mm) was packed
with 60/80 mesh arid-washed Chromusorb \V con ted with 10% Free
Fatly Acid Plui,.-(FKAIM.
Column •>'. An S-linil foilrtl ^l:i>» o'lumn (i.<). 'J mm) was parked
with .Ml/liO nif>h t'hrmuuMirli 101. Sfirridl care was lakrn to l',u>h
the column sullicienily with helium i-lO-"><) min) to rrmnve any re-
sidual air faun the |xu-kmi; Iwlore heat inc.
Method I. An ur£:mir i»iu|Hiuiid tl «1) WHS va|mrizvd into the
rt-srrvtMr MIH] ulU'^iii lu lc;ik ultiwly ihriiuxh a ^laxs Irit
directly into the ion source of the mass spectrometer. The mass
spectrum of the compound was acquired repetitively trum 20 to
200 amu using computer control, and the data were stored on a
disk storage device. As the MS data were being generated and
stored. 1 »d of tap water was injected onto GC Column 1 at 100°.
After the water eluted from the column into the detector and was
pumped out of the MS. valves were opened and the organic com-
pound was pumped from the batch inlet. The experiment was ter-
minated, and the data were recalled for evaluation of the mass
spectra. The organic compounds that were analyzed using this
method include the following: 1.2-dimethoxvethane (DMK). di-n-
butyl amine. svc-butyl alcohol, methyl-n-octanoate. acetic acid, n-
hexadecane. n-amyl alcohol, and n-butyl acetate.
Method 2. U'ater 11 n\) was vaporized into the batch inlet reser-
voir and allowed to leak continuously into the detector. Then 1 ul
of acetone containing 100 ng each of n-decane, methyl-IM.butyl ke-
tone (MIBK). n-butyl acetate, and /i-amyl alcohol was injected
onto Column 1 at 70°. After the ncetone solvent was pumped out
of the detector, the ionizer was turned on and the mn.ss spectra of
the organic compounds were repetitively scanned whilo water was
continuously leaking into the ion source. The GC temperature was
programmed from 70 to IVJO" jit 6°/min.
Method 3. The m.»s spt-ctra ul selected organic compounds
were measured liy injcv'tm" aquivnis :iiul/ur.ai-rtiiiu* >«>liili\in» into
the (iC/MS. The roni|hiumls sturiifd. i|iiantilu-s. Milvrnls. .u»l (iC
conditions arc shown in Tahlc 1.
Waste Effluent Sample. This s;impl«- w;ts n<-()uin-(l from the ef-
llurnt of a la>:iN>ii typ«- flu-mical trcaliiu-nl l.u-ility and '.vj> ana-
lyzed by direct.injectiun ul 1 n\ IMIIO i'uluinn '.i. The nilumn tcm-
ANALYTICAl. CHEMISTRY. VOL. 46. NO. 13. NOVEMBER 1974 • 1913
-------
Butyl Act late
From me
Batch Inlet
DT—sj
Background
in System
18 23 33 KJ S3 60 78 83 33 100 118 123 130 110
SflECTH/l H/BEH
Figure 1. An ion aburviance ctvomatogram obtained from water eluting from Column 1 while n-butyl acetate was permitted to leak continuous-
ly into the ion source from the batch inlet
perature was held at 120* for 3 min, then programmed to 180° at
l2Vmin.
Tap Water Sample. A 5-«l sample of tap water was injected
onto Column 3 at 140°. Ions of mass 47, S3. and So were repetitive-
ly observed with an integration time of 450. 900, or 1350 msec
which was determined dynamically to maximize signal to noise.
The technique of subset data acquisition was described previously
(2). The technique of integration time a» a function of signal
strength will be described in the near future ie>. Additional details
on the analysis of chloroform in tap water will be presented in a
future paper (9).
RESULTS AND DISCUSSION
The mass spectra of a number of organic compounds
were recorded while water was present in relatively high
proportion in the ion source. The purpose of these experi-
ments was to determine the effects of water on well known
fragmentation patterns. Method 1 was used to study the ef-
fect of an increase in water pressure from the GC inlet
(Column 1) on the fragmentation patterns of several com-
pounds which were permitted to leak continuously from
the batch inlet. For example, mass spectra of n-butyl ace-
tate were acquired and recorded continuously during one
experiment and Figure 1 shows the ion abundance chro-
matogram that was obtained. In regions I, II, and III, n-
butyl acetnte spectra were acquired. In region II, water
eluted from the column and subsequently was pumped out
of the mass spectrometer. .In region IV, the batch inlet sys-
tem valve w.i$ closed and n-butyl acetate wns pumped from
the mass spectrometer. From this experiment, mass spectra
of n-butyl acetate were plotted before (spectrum 60), dur-
ing (spectrum 75), and after (spectrum So) water eluted
from the GC inlet system. Spectrum 130 was subtracted
from each spectrum to correct for background ions. The
masses and abundance data from these spectra are shown
in Table II. For any given ion, the relative abundance is
netrly identical in all three spectra, which indicated that
xvnicr had no significant effect on the fragmentation pro-
ct-sses of this compound.
(8) J W. Eicfvritiergof. 1. E. Harris, and W. U Budda. presented at trie 22nd
annual c.-nlci^ce on mass spoclronwuy 0nd >lto>d loptcx. Pnrtadetphia.
Pa.. May 13-24. 1974. In press.
(9) T. A. Bdi.ir and J J. Ucriteruxxg. Environ. Sd Tocfinot.. In pr«sa.
Table II. Mass Spectra of Butyl Acetate from Batch
Inlet with Water Eluting from GC
Relative ion abundance*
m/«
39
40
41
42
43
44
45
55
56
57
58
61
71
73
87
Spectrum
No. 00-130
5
1
17
5
100
. 3
2
7
35
8
2
12
2
12
2
Spectnsn
So. 75-130
5
0
19
4
100
4
2
7
32
8
2
12
2
12
2
Spectrum
JCoVSS-130
6
1
19
6
100
3
2
7
36
9
2
13
2
14
2
Similar results were obtained with the other compounds
studied by Method 1. These compounds included methyl
n-octanoate which has a base peak in its mass spectrum
that is due to a McLalTerty rearrangement. In this well
known process, a gamma hydrogen is transferred to the
carbunyl oxygen and a neutral Cr.Hia fragment is expelled
from the molecular ion. It was clear that the relatively high
pressure of water present during the ionizalion did not
cause ion-molecule reactions or other effects that would
alter this fragmentation process. We concluded that the ion
source design, pumping speeds, etc. that were employed
were such that disruptive effects were precluded in general.
This conclusion was confirmed by several experiments
which used Method 2 to study the effect of a constant
water pressure from the batch inlet on the fragmentation
p.ittern of several compounds introduced from the ( H«-U/C»JOCO
-------
!„
i.
t.
f.
r
5*.
*..
K.
«.
t.
u-4
ISMKt
1011
; I
I •
f» !• US «i B» IB to <••
figure 2. An ion abundance crtromatogram Obtained from a mixture of SO ng of each of the following compounds in tap wafir on Column 2
(1) carton dioxide tram the tap water sotwenc (2) etnyl acetate: (3) 1.2-dimetnoxyethane: (4) methyl isobutyt ketone: (S) Ooxane: (6) waier [7> •cetoptenone: (8)
»cMorophenot: and (9) /rvcrtsol
time the compounds were undergoing ionization and frag-
mentation. One of the compounds studied was n-amyl alco-
hol which undergoes an electron impact induced dehydra-
tion reaction. In this and all other experiments the pres-
-nce of water in the MS detector system caused no notice-
able effect on the observed mass spectra.
A number of additional experiments (Method 3) were
conducted to support these conclusions. The compounds
shown in Table I were introduced into the mass spectrome-
ter through the GC inlet. The compounds were selected as
representatives of several classes of compounds commonly
found as voiaciles in water samples. The mass spectra that
were obtained from acetone and water solutions were com-
pared and found to be identical in ail cases.
Precautions and System Performance. In principle,
the sample enrichment device in the GC/MS interface
should reduce the amount of water which enters the ion
source of the mass spectrometer. Although no quantitative
measurements were made, our qualitative observations
support this expectation. As much as 10 u.1 of water was in-
troduced onto the GC column in a single injection and a
number of 1- to 10-»il injections were made during a work
day with no apparent detrimental effects on system perfor-
mance or sensitivity.
With cross-linked porous polymer packed columns, e.g.,
column 3, from which water elutes very quickly, the ion
source potentials (5-100 V) and electron multiplier voltage
(3 kV) were not applied and data acquisition was not begun
until water eluted. This is also standard procedure tor or-
ganic solvent extracts. With other columns, e.g.. Columns 1
and 2. from which water eiutes much later and after some
organics. the source and multiplier potentials were applied
immediately after injection and left on during solvent eiu-
tion with small, i.e., 1 n\. water injections. With larger
quantities, these potentials were usually removed during
elution of the water (Figure 2). A downward drift in the
ionizing current was observed with the ion source on while
1 ill or more of water eiuted (Fiirure 1).
During the course of these experiments over more than
IS months, frequent observations were made of overall sys-
tem performance. The jx>rformance measurement was the
ability of the system to produce, from a '_'b-ng injection, the
furrifrt electron impact fragmentation pattern of a refer-
ence compound US) with u molecular weight of -442. The
background inn's* was never observed at jrre-Jter than '2-;J"»»
of the base peak. Normal degradation in the performance
nf a qiKiiirupole mass .spectrometer is revealed by a loss in
M-iisitivity (signal/noise) and resolution ut musses greater
than •—2.~»0 amu. This is caused frequently by an accumula-
tion of carbon and other extraneous deposits on the ion
source and i|u:uirupole rods. Surprisingly, it was our quali-
tative observation that normal performance degradation
was retarded somewhat during the period of intensive
study of aqueous injections. We tentatively credit this ap-
parent effect to a steam cleaning phenomena.
Background Subtraction. Care must be taken during
computer assisted background subtraction if the well es-
tablished fragmentation pattern? are to be observed. For
example, n-decane was chromatographrd in acetone and
water on Column 1 (Table I). The base peak of n-decane.
after background subtraction, in acetone wa* mass 57 and,
in water, mass 43. The background spectrum selected for
subtraction in each case was immediately !n front of the n-
decane peak. Acetone has a base peak of m;iss 43 and resid-
ual acetone in the spectrometer contributed a large mass 43
ion to the background spectrum. Therefore, the mass 43
ion abundance was reduced substantially by background
subtraction in the spectrum of n-decane in acetone, and
mass 57 became the base peak. This did not occur in water
and is a clear advantage of a non-organic solvent for GC/
MS analyses of organic pollutants.
However, an ion of mass 44 was observed as the base
peak for a tap water blank on Column 1 (data acquisition
began at mass 20). This was not caused by dissolved carbon
dioxide, which generally elutes as a sharp peak (Figure 2).
It may emanate from the continuous decomposition of car-
bonates or other dissolved compounds, n-Butyl acetate in
water eiuted from Column 1 on the leading edge of the
water peak. Subtraction of a background spectrum from
before the n-butyl acetate peak save a spectrum similar to
that of n.-butyl acetate but with mass 44 as the base peak.
If a spectrum for subtraction was chosen from near the
apex of the water peak, the mass 44 was eliminated and the
correct n-butyl acetate spectrum was obtained. This illus-
trates one necessary precaution of aqueous injection GC/
MS. The ability to rapidly (10-15 sec) view a mass spec-
trum histogram on a CRT display greatly facilitates back-
ground subtraction and other type* of real time interactive
data reduction.
Column Selection. The selection of a GC column for
aqueous GC/MS analysis depends <>n the types of com-
pounds sought in the analysis. If it is desired to search for
low molecular weight volatile compounds, porous polymers
(e.g.. Column ;!) appear to be the best choice lor a column
packing material. If the aqueous analysis is to be extended
into the types of compounds usually sought !n- solvent ex-
traction, f.fi.. phenol and substituted phrnnls. .mother col-
umn is required because elution. tjtws lor higher molecular
\v«-i;jht compounds become unacceptable. Kiiher I'arhowax
(Column 1) ur KFA1* (Column 2) si.itionarv |ih:i*fs are a
reasonable choice. The disaiiviini.i^es ol these columns in-
clude the inability to observe, as JLstinot peaks, compounds
ANALYTICAL CHEMISTRY. VOL. 4& wn
1971 . .«..*
-------
1}
*'.
If.
Figure 3. An ion aOundance chromatogram obtained from a direct injection onto Column 3 of effluent from a lagoon chemical treatment facility
The compounds ocntihetf weft: (1) metfunot (2) etrunol: (3>acetone: (4) 2-vooanoi: (51 acetic acid; (6) 2-butanone: (7) propanoic acid: (8) 2-«moryetfiano<; (9)
. i tQi inetriyi armnc: iiii M.AM>n«(nylfermanrude, and (12) pentanoic acd
TR*
ira 110
Figure 4. ion aouncance cnromatogram obtained from tap water
ekmng from Column 3 witn suaset data acquisition at masses 47,
83. and 85
which have the same retention time as water. Also, as
pointed out previously, it may be necessary to interrupt
data acquisition curinz eiuuon of water with these columns
(Figure 2). None of the numerous specialty phases avail-
able for specific analyses were evaluated for this applica-
tion.
Application to an Environmental Sample. Numerous
water samples were collected trorn waste effluents, chemi-
cal spill.-, and waste treatment plants and submitted to
aqueous inject inn CiC MS. The inn abundance chromato-
gram shown in Figure •'? wa.- uhtained I'rtim an aqueous in-
jection of the effluent from a lagoon type chemical waste
treatment plant. From each peak of the chromatogrnm. a
mas.s spectrum \va> retrieved Irom the disk storage device.
This provided an unequivocal identification of the pollu-
tants .-till present in the effluent.
Detection Limits. The absolute detection limit of a qua-
drupole CC/MS system will vary widely and depend on a
variety uf factors including the efficiency of the GC column
and enrichment device, the presence of carbon deposits on
the ion source or rods, the adjustment, of the ion source and
rod potentials, the mass range scanned, the integration
time per mass unit, the U>lal scan time, and the signal-lo-
noise ratio required in any given mass spectrum. During
the course of these experiments with the quadrupole GC/'
MS, it was possible to obtain a reasonably clean (signal/
noise = 25 or greater) 40—{00 amu mass spectrum in a 5-sec
total scan time from about 5 ng of a volatile organic com-
pound. A 5-jil aqueous solution containing a total of 5 ng of
a compound has a concentration of 1 mg/L (1 ppm) and this|
should be the approximate lower detection limit for a 40-
400 amu 5-sec scan. Using a somewhat shorter mass range.
(20-200 amu). methylene chloride and n-ainyl alcohol were
mixed in water at the 1 mg/L concentration and chromato-
graphed using Column 1 (Table I). Acceptable mass spectra
were obtained from this experiment. However,, with the po-
rous polymer packed column (Column 3), the detection
level was about 10-20 rag/1. In general, the concentration
required to obtain a reasonably clean MS was between 1-50
mg/L This detection limit for conventional data acquisition
is usually not sufficient for relatively clean water, i.e., fin-
ished tap wacer or surface waters. However, it is more than
adequate for the analysis of effluents and other relatively
dirty water which frequently contains organic compounds
at a concentration greater than 1 mg/1.
There are several methods available which may be used
to extend the detection limit. Very large samples, i.e., 50-
100 jil may be injected if a solvent venting valve is installed
and a column is used from which water eiutes rapidly, e.g.,
Column 3. Another approach to enhance sensitivity for
specific compounds utilized data acquisition from subsets
of the ions which are observed in conventional mass spec-
trometry (2). With this approach only a relatively few ions
are monitored in real time with a relatively long integration
time on each to substantially Improve signal/noise by time
averaging. Figure 4 shows a direct aqueous analysis of 5 n\
of a finished tap water sample. Ions of mass 47, S3. and 85
were repetitively monitored with an integration time of
450-1350 msec which was determined dynamically as a
function of signal strength. These ion* were .selected be-
cause they represent the three most abundant ions in the
mass spectrum of chloroform. The peak observed had the
retention time of chloroform which was estimated to be
present in the 80-120 jtg/1. concentration range;.
CONCLUSION
Direct aqueous injection - gas chromatogTaphy with a
computer controlled quadrupole mass spectrometer detec-
tor is a powerful supplemental method for the unambig-,
uoii£ identification of the more volatile organic pollutants
in water sample*. Hela.tively large pressures of water vapor
in the mass .spectrometer have no significant effect on the
-------
weii established 70-«V electron impact frn::montalion pat- This technique made the mHlwwl applicable to the analysis
terns of organic compounds or the perfurniancc of the qua- of relatively clean surface »r n'uikinp water.
drupoie GC/MS system. The detection limits attained
using conventional data acquisition wer«? l-SO ppm which Aownwi •-nr \TPMT
makes the technique compatihle vith the concentrations of AC*U>U»> i-«.i;u.ne.r« i
organic compounds found in domestic sewage and other We thank Torn Bellar of th:.« laboratory for the original
waste effluent water samples. Greater sensitivity, to about suggestion to analyze lap WHUV for chloroform.
50 ppb, was attained with real time data acquisition from
subsets of the ions used in conventional mass spectrometry. RKCKlVF.lt for review May 8. 1974. Accepted July 19,1974.
-------
CE NO. 6
DETERMINATION OF ORGANOCHLORINE PESTICIDES IN INDUSTRIAL EFFLUENTS
-------
1. METHOD FOR ORGANOCHLORINE PESTICIDES IN INDUSTRIAL EFFLUENTS
1. Scope and Application
1.1 This method covers the determination of various organochlorine
pesticides, including some pesticidal degradation products and related
compounds in industrial effluents. Such compounds are composed of
carbon, hydrogen, and chlorine, but may also contain oxygen, sulfur,
phosphorus, nitrogen or other halogens.
1.2 The following compounds may be determined individually by this method
with a sensitivity of 1 ug/liter: BHC, lindane, heptachlor, aldrin,
M
I |
heptachlor epoxide, dieldrin, endrin, Captan, DDE, ODD, DDT, methoxy-
•
-p
P* chlor, endosulfan, dichloran, mirex, pentachloronitrobenzene and tri-
~ ^ fluralin. Under favorable circumstances, Strobane, toxaphene,
chlordane (tech.) and others may also be determined. The usefulness
of the method for other specific pesticides must be demonstrated by
-— *3 the analyst before any attempt is made to apply it to sample analysis.
•
;!.«? When organochlorine pesticides exist as complex mixtures, the
. individual compounds may be difficult to distinguish. High, low, or
P=i otherwise unreliable results may be obtained through misidentifica-
tion and/or one compound obscuring another of lesser concentration.
Provisions incorporated in this method are intended to minimize the
occurrence of such interferences.
Summary
-.1 The method offers several analytical alternatives, dependent on the
analyst's assessment of the nature and extent of interferences and/or
the complexity of the pesticide mixtures found. Specifically, the
procedure describes the use of an effective co-solvent for efficient
sample extraction; provides, through use of column chromatography
~ a
<
_ "co"
^S,
-------
1-2
ana liquid-liquid partition, methods for elimination of non-pesticide
interferences and the pre-separation of pesticide mixtures. Identifi-
cation is made by selective gas chromatographic separations and may
be corroborated through the use of two or more unlike columns.
Detection and measurement is accomplished by electron capture, micro-
couloraetric or electrolytic conductivity gas chromatography. Results
are reported in micrograms per liter.
2.2 This method is recommended for use only by experienced pesticide
analysts or under the close supervision of such qualified persons.
3. Interferences
3.1 Solvents, reagents, glassware, and other sample processing hardware
may yield discrete artifacts and/or elevated baselines causing
misinterpretation of gas chromatograms. All of these materials must
be demonstrated to be free from interferences under the conditions
of the analysis. Specific selection of reagents and purification of
solvents by distillation in all-glass systems may be required.
Refer to Part I, Sections 1.4 and 1.5, (1).
3.2 The interferences in industrial effluents are high and varied and
often pose great difficulty in obtaining accurate and precise
measurement of organochlorine pesticides. Sample clean-up procedures
are generally required and may result in the loss of certain organo-
chlocine pesticides. Therefore, great care should be exercised in
the selection and use of methods for eliminating or minimizing,
interferences. It is not possible to describe procedures for over-
coming all of the interferences that may be encountered in industrial
effluents.
-------
1-3
3.3 Pol/chlorinated Biphenyls (PCB's) - Special attention is called
to industrial plasticizers and hydraulic fluids such as the PCB's
which are a potential source of interference in pesticide analysis.
The presence of PCB's is indicated by a large number of partially
resolved or unresolved peaks which may occur throughout the entire
chromatogram. Particularly severe PCB interference will require
special separation procedures (2,3).
3.4 Phthalate Esters - These compounds, widely used as plasticizers,
respond to the electron capture detector and are a source of inter-
ference in the determination of organochlorine pesticides using
this detector. Water leaches these materials from plastics, such
as polyethylene bottles and tygon tubing. The presence of phthalate
esters is implicated in samples that respond to electron capture but
not to the microcoulometric or electrolytic conductivity halogen
detectors or to the flame photometric detector.
5.5 Organophosphorus Pesticides - A number of organophosphorus pesticides,
such as those containing a nitro group, eg, parathion, also respond
to the electron capture detector and may interfere with the determina-
tion of the organochlorine pesticides. Such compounds can be
identified by their response to the flame photometric detector
4. Apparatus and Materials
4.1 Gas Chroraatograph - Equipped with glass lined injection port.
4.2. Detector Options:
4.2.1 Electron Capture - Radioactive (tritium or nickel 63)
4.2.2 Microcoulometric Titration
4.2.5 Electrolytic Conductivity
-------
1- 4
4.3 Recorder - Potentiometric strip chart (10 in.) compatible with
the detector.
4.4 Gas Chromatographic Column Materials:
4.4.1 Tubing - Pyrex (180 cm long x 4 ram ID)
4.4.2 Glass Wool - Silanized
4.4.3 Solid Support - Gas-Chrom Q (100-120 mesh)
4.4.4 Liquid Phases - Expressed as weight percent coated on
solid support.
4-.4.4.1 OV-1, 3%
4.4.4.2 OV-210, 5%
4.4.4.3 OV-17, 1.5% plus QF-1, 1.95%
4.4.4.4 QF-1, 6% plus SE-30, 4%
4.5 Kuderna-Danish (K-D) Glassware (Kontes)
4.5.1 Snyder Column - three ball (macro) and two ball (aicro)
4.5.2 Evaporative Flasks - 500 ml
4.5.5 Receiver Ampuls - 10 ml,graduated
4.5.4 Ampul Stoppers
4.6 Chromatographic Column - Chromaflex (400 mm long x 19 mm ID) with
coarse fritted plate on bottom and Teflon stopcock; 250 ml reservoir
bulb at top of column with flared out funnel shape at top of bulb - a
special order (Kontes K-420540-9011).
4.7 Chromatographic Column - pyrex (approximately 400 mm long, x 20 mm ID)
with coarse fritted plate on bottom.
4.3 Micro Syringes - 10, 25, 50 and 100 ul
4.9 Separatory Funnels - 125 ml, 1000 ml and 2000 ml with Teflon stopcock.
4.10 Blender - High speed, glass or stainless steel cup.
-------
1-5
4.11 Graduated cylinders - 100 and 250 ml
4.12 Florisil - PR Grade (60-100 mesh); purchase activated at 1250 F
and store in the dark in glass containers with glass stoppers or
foil-lined screw caps. Before use, activate each batch overnight
at 130 C in foil-covered glass container. Determine lauric-acid
value (See Appendix I).
5. Reagents, Solvents, and Standards
5.1 Ferrous Sulfate - (ACS) 30% solution in distilled water.
*
5.2 Potassium Iodide - (ACS) 10% solution in distilled water.
5.3 Sodium Chloride - (ACS) Saturated solution in distilled water
(pre-rinse NaCl with hexane).
5.4 Sodium Hydroxide - (ACS) 10 N in distilled water.
5.5 Sodium Sulfate - (ACS) Granular, anhydrous (conditioned @ 400 C for 4 hrs).
5.6 Sulfuric Acid - (ACS) Mix equal volumes of cone. H.SO. with
distilled water.
5.7 Diethyl Ether - Nanograde, redistilled in glass, if necessary.
5.7.1 Must contain 2% alcohol and be free of peroxides by
following test: To 10 ml of ether in glass-stoppered
cylinder previously rinsed with ether, add one ml of
freshly prepared 10% KI solution. Shake and let stand
one minute. No yellow color should be observed in either layer.
5.7.2 Decompose ether peroxides by adding 40 g of 30% ferrous sulfate
solution to each liter of solvent. CAUTION: Reaction maybe
vigorous if the solvent contains a high concentration of
peroxides.
5.7.3 Distill deperoxidized ether in glass and add 2% ethanol.
-------
1-6
5.8 Acetonitrile, Heoane, Methanol, Methylene Chloride, Petroleum
Ether (boiling range 30-60 C) - nanograde, redistill in glass
if necessary
5.9 Pesticide Standards - Reference grade.
6. Calibration :
6.1 Gas chromatographic operating conditions are considered acceptable
if the response to dicapthon is at least 50% of full scale when
< 0.06 ng is injected for electron capture detection and < 100 ng is
injected for microcoulometric or electrolytic conductivity detection.
For all quantitative measurements, the detector must be operated
within its linear response range and the detector noise level should
be less than 2% of full scale.
6.2 Standards are injected frequently as a check on the stability of
operating conditions. Gas chromatograms of several standard
pesticides are shown in Figures 1, 2, 3 and 4 and provide reference
operating conditions for the four recommended columns.
6.3 The elution order and retention ratios of various organochlorine
pesticides are provided in Table 1, as a guide.
7- Quality Control
7.1 Duplicate and spiked sample analyses are recommended as quality cpntrol
checks. When the routine occurrence of a pesticide is being observed,
the use of quality control charts is recommended (5).
7.2 Each time a set of samples is extracted, a method blank is determined
on a volume of distilled water equivalent to that used to dilute the
sample. •
-------
1-7
8. Sample Preparation
8.1 Blend the sample if suspended matter is present and adjust pH to
near neutral (pH 6.5-7.5) with 50% sulfuric acid or 10 N sodium
hydroxide.
8.2 For a sensitivity requirement of 1 yg/1, when using microcoulometric
or electrolytic conductivity methods for detection, 100 ml or more
of sample will be required for analysis. If interferences pose no
problem, the sensitivity of the electron capture detector should
permit as little as 50 ml of sample to be used. Background informa-
tion on the extent and nature of interferences will assist the
analyst in choosing the required sample size and preferred detector.
8.3 Quantitatively transfer the proper aliquot into a two-liter
separatory funnel and dilute to one liter.
9. Extraction
9.1 Add 60 ml of 15% methylene chloride in hexane (v:v) to the sample
in the separatory funnel and shake vigorously for two minutes.
9.2 Allow the mixed solvent to separate from the sample, then draw the
water into a one-liter Erlenmeyer flask. Pour the organic layer
into a 100 ml beaker and then pass it through a column containing
3-4 inches of anhydrous sodium sulfate, and collect it in a 500 ml
K-D flask equipped vith a 10 ml ampul. Return the water phase to
the separatory funnel. Rinse the Erlenmeyer flask with a second
60 ml volume of solvent; add the solvent to the separatory funnel
and complete the extraction procedure a second time. Perform a
third extraction in the same manner.
9.3 Concentrate the extract in the K-D evaporator on a hot water bath.
-------
1-8
9.4 Analyze by gas chromatography unless a need for cleanup is indicated.
(See Section 10).
10. Clean-up and.Separation Procedures
10.1 Interferences in the form of distinct peaks and/or high background
in the initial gas chromatographic analysis, as well as the physical
characteristics of the extract (color, cloudiness, viscosity) and
background knowledge of the sample will indicate whether clean-up
is required. When these interfere with measurement of the pesticides,
or affect column life or detector sensitivity, proceed as directed
below.
10.2 Acetonitrile Partition - This procedure is used to isolate fats and
oils from the sample extracts. It should be noted that not all
pesticides are quantitatively recovered by this procedure. The
analyst must be aware of this and demonstrate the efficiency of
the partitioning for specific pesticides. Of the pesticides listed
in Scope (1.2) only mirex is not efficiently recovered.
10.2.1 Quantitatively transfer the previously concentrated extract
, to a 125 ml separatory funnel with enough hexane to bring
the final volume to 15 ml. Extract the sample four times
by shaking vigorously for one minute with 30 ml portions
of hexane-saturated acetonitrile. :
10.2.2 Combine and transfar the acetonitrile phases to a one-liter
separatory funnel and add 650 ml of distilled water and
40 ml of saturated sodium chloride solution. Mix thoroughly
for 30-45 seconds. Extract with two 100 ml portions of
-------
1-9
hexane by vigorously shaking about 15 seconds.
10.2.3 Combine the hexane extracts in a one-liter separatory funnel
and wash with two 100 ml portions of distilled water. Dis-
card the water layer and pour the hexane layer through a
3-4 inch anhydrous sodium sulfate column into a 500 ml K-D
flask equipped with a 10 ml ampul. Rinse the separatory
funnel and column with three 10 ml portions of hexane.
10.2.4 Concentrate the extracts to 6-10 ml in the K-D evaporator
in a hot water bath.
10.2.5 Analyze by gas chromatography unless a need for further
cleanup is indicated.
10.3 Florisil Column Adsorption Chromatography
10.3.1 Adjust the sample extract volume to 10 ml.
10.3.2 Place a charge of activated Florisil (weight determined by
lauric-acid value, see Appendix I) in a Chromaflex column.
After settling the Florisil by tapping the column, add about
one-half inch layer of anhydrous granular sodium sulfate to
the top.
10.3.3 Pre-elute the column, after cooling, with 50-60 ml of
petroleum ether. Discard the eluate and just prior to
exposure of the sulfate layer to air, quantitatively transfer
the sample extract into the column by decantation and subse-
quent petroleum ether washings. Adjust the elution rate to
about 5 ml per minute and, separately, collect up to three
eluates in 500 ml K-D flasks equipped with 10 ml ampuls.
(See Eluate Composition 10.4).
-------
1-10
I.
Perfcra: the first s?U:C-. on ». •. :>. *\V :n'. .-•: f<% rthvl other in
petroleum ether, and the second elution with 200 ml of 15%
ethyl ether in petroleum ether. Perform the third elution
with 200 ml of 50% ethyl ether - petroleum ether and the
fourth elution with 200 ml of 100% ethyl ether.
10.3.4 Concentrate the eluates to 6-10 ml in the K-D evaporator
in a hot water bath.
10.3.5 Analyze by gas chromatography.
i
10.4 Eluate Composition - By using an equivalent quantity of any batch of
Florisil as determined by its lauric acid value, the pesticides will
be separated into the eluates indicated below:
6% Eluate
Aldrin DDT Pentachloro-
BHC Heptachlor nitrobenzene
Chlordane Heptachlor Epoxide Strobane
ODD Lindane Toxaphene
DDE Methoxychlor Trifluralin
Mirex PCB's
15% Eluate 50% Eluate
Endosulfan I Endosulfan II
Endrin Captan
Dieldrin
Dichloran
Phthalate esters
Certain thiophosphate pesticides will occur in each of the above
fractions as well as the 100% fraction. For additional information
/
regarding eluate composition, refer to the FDA Pesticide Analytical
c
Manual (6).
-------
1-11
11. Calculation of Results
11.1 Determine the pesticide concentration by using the absolute calibra-
tion procedure described below or the relative calibration procedure
described in Part I, Section 3.4.2. (1).
(1) Micrograms/liter = (A) (B) (Vt)
(Vj) (Vs)
A = ng standard
Standard area
B » Sample aliquot area
V. = Volume of extract injected (ul)
i
V = Volume of total extract (yl)
V = Volume of water extracted (ml)
12. Reporting Results
12.1 Report results in micrograms per liter without correction for
recovery data. When duplicate and spiked samples are analyzed,all
data obtained should be reported.
-------
1-12
REFERENCES
1. "Method for Organic Pesticides in Water and Wastewater," Environmental
Protection Agency, National Environmental Research Center, Cincinnati, Ohio
45268, 1971.
2. Monsanto Methodology for Aroclors - Analysis of Environmental Materials for
Biphenyls, Analytical Chemistry Method 71-35, Monsanto Company, St. Louis,
Missouri 63166, 1970.
3. "Method for Polychlorinated Biphenyls in Industrial Effluents," Environmental
Protection Agency, National Environmental Research Center, Cincinnati, Ohio
45268, 1973.
4. "Method for Organophosphorus Pesticides in Industrial Effluents," Environ-
mental Protection Agency, National Environmental Research Center, Cincinnati
Ohio 45268, 1973..
5. "Handbook for Analytical Quality Control in Water and Wastewater Laboratories,"
Chapter 6, Section 6.4, U.S. Environmental Protection Agency, National Environ-
mental Research Center, Analytical Quality Control Laboratory, Cincinnati,
Ohio 4S268, 1973.
6. "Pesticide Analytical Manual," U.S. Dept. of Health, Education and Welfare,
Fqod and Drug Administration, Washington, D,C.
7. "Analysis of Pesticide Residues in Human and Environmental Samples," U.S.
Environmental Protection Agency, Perrine Primate Research Laboratories,
Perrine, Florida 33157, 1971.
8. Mills, P.A., "Variation of Florisil Activity: Simple Method for Measuring
Adsorbent Capacity and its Use in Standardizing Florisil Columns," Journal
of the Association of Official Analytical Chemists, 51, 29 (1968).
9. Goerlitz, D.F. and Brown, E., "Methods for Analysis of Organic Substances
in Water," Techniques of Water Resources Investigations of the United States
Geological Survey, Book 5, Chapter A3, U.S. Department of the Interior,
Geological Survey, Washington, D.C. 20402, 1972, pp. 24-40.
i
10. Steere, N.V., editor, "Handbook of Laboratory Safety," Chemical Rubber
Company, 18901 Cranwood Parkway, Cleveland, Ohio 44128, 1971, pp. 250-254.
-------
1-15
Table 1
RETENTION RATIOS OF VARIOUS ORGANOCHLORINE PESTICIDES RELATIVE TO ALDRIN
Liquid Phase
Column Temp.
Argon/Methane
Carrier Flow
Pesticide
Trifluralin
«-BHC
PCNB
Lindane
Dichloran
Heptachlor
Aldrin
Heptachlor Epoxide
Endosulfan I
p,p'-DDE
Dieldrin
Cap tan
Endrin
o,p'-DDT
p,p'-DDD
Endosulfan II
p,p'-DDT
Mi rex
Methoxychlor
Aldrin
(Min absolute)
1.5% OV-17
+
1.95% QF-1
200 C
60 ml/min
RR
0.39
0.54
0.68
0.69
0.77
0.82
1.00
1.54
1.95
2.23
2.40
2.59
2.93
3.16
3.48
3.59
4.18
6.1
7.6
3.5
5%
OV-210
180 C
70 ml/min
RR
1.11
0.64
0.85
0.81
1.29
0.87
1.00
1.93
2.48
2.10
3.00
4.09
3.56
2.70
3.75
4.59
4.07
3.78
6.5
2.6
3%
OV-1 ,
180 C
70 ml/min
RR
0.33
0.35
0.49
0.44
0.49
0.78
1.00
1.28
1.62
2.00
1.93
1.22
2.18
2.69
2.61
2.25
3.50
6.6
5.7
4.0
6% QF-1
+
4% SE-30
200 C
60 ml/min
RR
0.57
0.49
0.63
0.60
0.70
0.83
1.00
1.43
1.79
1.82
2.12
1.94
2.42
2.39
2.55
2.72
3.12
4.79
4.60
5.6
All columns glass, 180 cm I 4 mm ID, solid support Gas-Chrom Q (100/120 mesh)
-------
1-1
•\rn-\PT\ \
13. Standardization of Florisil Column by Weight Adjustment Bas^d on Ad
o£ Laurie Acid.
13.1 A rapid method for determining adsorptive capacity of Florisil is
based on adsorption of lauric acid from hexane solution (6) (8).
An excess of lauric acid is used and amount not adsorbed is measured
by alkali titration. Weight of lauric acid adsorbed is used to
calculate, by simple proportion, equivalent quantities of Florisil
for batches having different adsorptive capacities.
13.2 Apparatus
13.2.1 Buret. -- 25 ml with 1/10 ml graduations.
13.2.2 Erlenmeyer flasks. — 125 ml narrow mouth and 25 ml, glass
stoppered.
13.2.3 Pipet. -- 10 and 20 ml transfer.
13.2.4 Volumetric flasks. -- 500 ml.
13.3 Reagents and Solvents
t
13.3.1 Alcohol, ethyl. — USP or absolute, neutralized to
phenolphthalein.
13.3.2 Hexane. — Distilled from all glass apparatus.
15.3.3 Lauric acid. —Purified, CP.
13.5.4 Lauric acid solution. -- Transfer 10.000 g lauric acid to
500 ml volumetric flask, dissolve in hexane, and dilute to
500 ml (1 ml = 20 mg).
13.3.5 Pfafinolphthalein Indicator. -- Dissolve 1 g in alcohol and
dilute to 100 ml.
-------
1-2
13.3.6 Sodium hydroxide. -- Dissolve 20 g NaOH (pellets, reagent
grade) in water and dilute to 500 ml (IN). Dilute 25 ml
1N_ NaOH to 500 ml with water (0.05N). Standardize as follows:
Weigh 100-200 mg lauric acid into 125 ml Erlenmeyer flask.
Add 50 ml neutralized ethyl alcohol and 3 drops phenol-
phthalein indicator; titrate to permanent end point. Calculate
mg lauric acid/ml 0.05 N_ NaOH (about 10 mg/ml).
•
13.4 Procedure
13.4.1 Transfer 2.000 g Florisil to 25 ml glass stoppered Erlenmeyer
flasks. Cover loosely with aluminum foil and heat overnight
at 130°C. Stopper, cool to room temperature, add 20.0 ml
lauric acid solution (400 mg), stopper, and shake occasionally
for 15 min. Let adsorbent settle and pipet 10.0 ml of
supernatant into 125 ml Erlenmeyer flask. Avoid inclusion
of any Florisil.
13.4.2 Add 50 ml neutral alcohol and 3 drops indicator solution;
titrate with 0.05N_ to a permanent end point.
15.5 Calculation of Lauric Acid Value and Adjustment of Column Weight
13.5.1 Calculate amount of lauric acid adsorbed on Florisil as
follows:
Lauric Acid value = mg lauric acid/g Florisil = 200 - (ml
required for titration X mg lauric acid/ml 0.05N_NaOH).
13.5.2 To obtain an equivalent quantity of any batch of Florisil,
divide 110 by lauric acid value for that batch and multiply
by 20 g. Veri ;"v proper elution of pesticides by 13.6.
-------
1-3
13.6 Test for Proper Elution Pattern and Recovery of Pesticides:
Prepare a test mixture containing aldrin, heptachlor epoxide,
p,p'-DDE, dieldrin, Parathion and malathion. Dieldrin and
Parathion should elute in the 15% eluate; all but a trace of
malathion in the 50% eluate and the others in the 6% eluate.
-------
T
25
20
15 10
RETENTION TIME IN MINUTES
Figure 1. Column Packing: 1.5% OV-17 + 1.95% QF-1, Carrier Gas: Argon/Methane at 60 ml/min,
Column Temperature: 200 C, Detector: Electron Capture.
-------
15
10 5 0
RETENTION TIME IN MINUTES
Figure 2. Column Packing: 5% OV-210, Carrier Gas: Argon/Methans
at 70 ml/min, Column Temperature: 180 C, Detector:
Electron Capture.
-------
I I I
I I I
25
20
15 10
RETENTION TIME IN MINUTES
Figure 3. Column Packing: 6% QF-1 + 4% SE-30, Carrier Gas: Argon/Methane at 60 ml/min,
Column Temperature: 200 C, Detector: Electron Capture.
-------
3 •
S
3
25
20
5
15 10
RETENTION TIME IN MINUTES
Figure 4. Column Packing: 3% OV-1, Carrier Gas: Argon/Methane at 70 ml/min,
Column Temperature: 180 C, Detector: Electron Capture.
-------
NO. 7
METHODS FOR CHEMICAL ANALYSIS OF WATER AND WASTES
-------
CE NO. 8
DETERMINING SELENIUM
-------
Reprinted from Atomic Absorption Newsletter, Volume 14, No. 5
DETERMINING SELENIUM IN WATER, WASTEWATER, SEDIMENT, AND
SLUDGE BY FLAMELESS ATOMIC ABSORPTION SPECTROSCOPY
Theodore D. Martin* and John F. Kopp
Environmental Protection Agency
Environmental Monitoring & Support Laboratory
Cincinnati, Ohio 45268
and
Richard D. Ediger
The Perkin-Elmer Corporation
Lombard, Illinois 60148
ABSTRACT
A method has been developed for the
determination of selenium in freshwater,
wastewater, sediment, and sludge by
flamelese atomic absorption spectros-
copy. A simple and rapid sample prep-
aration is given with a description of the
interferences that affect the analysis.
Nickel nitrate is added to both standards
and samples to prevent losses by volatil-
ization. The method demonstrates good
precision with day-to-day variation of
the absorbance values at the 0.25 and 2.5
nanogram level (25 nl of 10 and 100 ng
Se/1) varying only ±11.6% and ±4.1%
respectively, at the 95% confidence level.
The sensitivity of the method is 20 pico-
grams which for many tap, surface, and
well waters extends the detection limit
to 0.2 ug Se/1 without the use of scale
expansion.
INTRODUCTION
RESUME
On a develop pe une methode pour la
determination du selenium dans des eaux
fraiches et residuelles, des sediments et
des boues par spectroscopie d'absorp-
tion atomique sans flarame. On donne
une methode simple et rapide pour la
preparation des echantillons ainsi
qu'une description des interferences qui
affectent 1'analyse.
On fait un ajout de nitrate de nickel aux
standards et aux echantillons de maniere
a empecher des pertes par volatilisation.
La methode offre une bonne precision
avec une variation des valeurs d'absorb-
ance de jour en jour de ± 11,6% et
± 4,1% seulement au seuil de confiance
de 95% et ce respectivement pour des
valeurs de 0,25 et 2,5 nanogrammes (25
jil de 10 et 100 ng Se/1). La sensibilite de
la methode est de 20 picogrammes, ce
qui pour de nombreuses eaux de distri-
bution, de surface et de puits donne une
detection limite de 0,2 ng Se/1, sans
usage d'expansion d'echelle.
ZUSAMMENFA5SUNG
Eine Methode zur Bestimmung von Selen
in Frischwasser, Abwasser, Sedimenten
und Klarschlamm wurde mittels der
Sammenlosen Atomabsorptions - Spek-
troskopie entwickelt. Es wird auf eine
einfache und rasche Probenaufbereitung
hingewiesen und die, die Analyse beein-
flussenden Interferenzen angegeben.
Nickelnitrat wird sowohl zu den Stan-
dards alsauch zu den Proben zugesetzt,
um eventuelle Verluste durch Fliichtig-
keit zu vermeiden. Die Methode zeigt
fiir Extinktionswerte beim 0,25 und 24
ng-Niveau (25 nl von 10 und 100 ng
Se/1) eine gute Prazision und Reprodu-
zierbarkeit, mit Abweichungen von nur
±11,6% beziehungsweise ±4,1%, bei
einer 95%igen Sicherheit. Die Empfind-
lichkeit der Methode von 20 pg erweitert
die Nachweisgrenze bis 0,2 ng Se/1 fiir
viele Leitungs.-0berflachen.-und Quel-
lenwasser, ohne der Anwendung einer
Skalendehnung.
The analytical determination of selenium has long been
a problem to the analytical chemist. It is similar to arsenic
in toxicity and reactivity yet is probably much more diffi-
cult to detect and measure. The procedure given in Stan-
dard Methods for the Examination of Water and Waste-
water, 13th Ed., 1971, is time-consuming, subject to many
interferences, and relatively insensitive thus requiring a
large volume of sample. Therefore, it is often omitted from
routine analysis. Additionally, the colorimetric reagent
often used I diaminobenzidine) has been placed on the pos-
sible carcinogenic listing.
The selenium concentration of most finished waters is
less than 10 ug, 1. However, the use of selenium in industry
is growing. A major use of selenium is in the glass industry
to color glass a deep red and to neutralize iron color. Sele-
nium is known to be present in almost all types of paper.
Selenium may be present in soils both as selenite and sele-
nate. Thus, it is likely to be found in surface waters. Al-
though trace amounts of selenium have been shown to be
nutritionally beneficial in some animal diets, exposure to
higher concentrations produces toxic effects. There are also
some implications that selenium is a carcinogen.
In addition to the four valence states in which selenium
may exist, a variety of organo-selenium compounds is
known. Therefore, to ensure measurement of total sele-
nium, any method de%ased must include an oxidation step.
During this digestion, it is most important to maintain
oxidizing conditions. Inorganic selenium is not appreci-
* Author to whom correspondence should be addressed.
ably volatilized during digestion in a mixture of nitric and
perchloric acids (1, 2, 3, 4), except in the presence of such
a large excess of organic material that charring occurs (3,
4). In general, selenium may be lost from acid selenite
solutions by reducing but not by oxidizing agents.
In attempting to avoid volatilization losses, several meth-
ods effected dissolution of the sample through combustion
with oxygen in closed systems (1, 5, 6, 7). This can be done
either with a Schoniger combustion flask or in a Parr
bomb. Watkinson (8), after comparing wet oxidation with
the oxygen flask combustion, found no significant differ-
ence between the results of the two methods. He preferred
oxidation with nitric and perchloric acids. The use of
perchloric acid, however, is discouraged because of safety
reasons.
Since most natural waters and waste effluents contain
low concentrations of selenium, conventional atomic ab-
sorption has not been used for the analysis because of its
relatively poor sensitivity. Since it forms a hydride similar
to arsenic, several investigators have applied the arsine-
type procedure with subsequent introduction in an argon-
hydrogen flame to selenium (9). This technique has the
advantage of eliminating interference resulting from the
matrix effect and improving the detection limit. Many
problems, however, have been encountered in determining
selenium in domestic and industrial wastewaters particu-
larly the interference of organics, hizh copper concentra-
tions, and difficulty in forming the hydride.
With the advent of flameless atomization devices and
electrodelesa discharge lamps the analytical working ran^e
for many elements has been extended.
ATOMIC ABSORPTION NEWSLETTER
Vol. 14, No. 5. September-October 1975
AA-861
109
-------
Baird et al. (101. recently reported a flameless AAS
method for the determination of selenium in wastewater
employing a carbon rod analyzer. The obvious advantage
of this mode of analysis is the absence of the usual high
levels of flame background normally responsible for de-
creased sensitivity. Bairc did observe the need for a pre-
liminary digestion with nitric and perchloric acid to
oxidize organic material and solubiiize the selenium before
injections into the carbon rod. Replicate analyses of sam-
ples containing approximately 10 (ig of selenium per liter
gave a relative standard deviation of 6.8%. Because of the
safety factor, the author? of this paper prefer a digestion
step using a combination of nitric acid-hydrogen peroxide.
This combination allows the integrity of the sample to be
broken down and all of the selenium to be solubilized while
a condition of oxidation is maintained. Complete ashing
occurs during the charring step after the sample has been
injected into the furnace.
Because of its volatility, the possible loss of selenium has
been a point of concern in flameless analysis. In the devel-
opment of a furnace method for the analysis of selenium
in biological materials. Ediger (11. 12) determined that
the addition of nickel nitrate prior to the drying step pro-
duces a condition where high chairing temperatures
(1200° to 1500^ i can be tolerated without the .loss of
selenium. This condition facilitates complete ashing and
the removal of some matrix constituents which may cause
subsequent interference during the atomization. Welcher.
et al. (13 I have also demonstrated the stability of selenium
in the presence of nickel or other heavy metals in the deter-
mination of trace elements in high temperature alloys.
Recently. Henn (14'i demonstrated the same effect with
the use of molybdenum after separation of the selenium
from metallic interferences with a cation exchange resin.
This paper describes a method incorporating the nitric
acid-hydrogen peroxide digestion procedure, followed by
the addition of nickel nitrate and the use of the HGA-2000
Graphite Furnace in connection with a selenium EDL for
the determination of selenium in water, wastewater, sedi-
ments, and sludges while focusing on the problems en-
countered durins the analvsis.
EQUIPMENT
A Perkin-Elmer Model 503 atomic absorption spectro-
photometer equipped with a Perkin-Elmer Model HGA-
2000 Graphite Furnace, a Deuterium Background Cor-
rector, and a Perkin-Elmer selenium Electrodeless Dis-
charge Lamp (EDL) was used for the analysis. The spec-
trophotometer was operated in the peak read mode. A
Perkin-Elmer Model 056 recorder on 10 millivolt span was
used to record the absorbance signals. The Deuterium Back-
ground Corrector was used to compensate for non-specific
absorption using nitrogen at a flow rate of 2 liters.'min to
purge the optics. The selenium EDL was operated at 9
watts, with a wavelength setting of 196.0 nm and a spectral
slit width of 0.7 nm. All equipment requiring 120 volts was
operated on regulated voltage with the spectropbotometer
connected to a Stabiline saturable reactor autotransformer
voltage regulator to insure'stable power.
The HGA-2000 Graphite Furnace was programmed for
drying at 125QC (.with varying times depending on the vol-
ume of aliquot injected): 30-sec charring at 1500°C*:
and 10-sec atomization at 2700CC. Argon was used as the
furnace purge gas at a flow rate of 3 divisions, and the flow
was interrupted automatically during atomization.
Drying times of 20 sec were used for volumes of 5 and 10
ul: 30 sec for 25 ul; 50 sec for 50 ul: 65 sec for 75 ul: and
80 sec for 100 ul.
Eppendorf microliter pipets with disposable tips were
used to inject the samples into the furnace.
REAGENTS AND STANDARDS
Nitric Acid
(HN03), concentrated, ACS reagent grade, redistilled.
Hydrogen Peroxide
(H202), 307c, ACS reagent grade.
Standard Selenium Solution
A stock solution of 1000 mg Se/1 was prepared by dis-
solving 0.3453 grams of selenous acid (actual assay 9^.6%
HoSeOa) in 200 ml of deionized distilled water. Diluie
working standards (1, 2, 5,10,40, 50 and 100 pig Se/1 were
prepared from a diluted stock solution of 10 mg Se/1 by
withdrawing the appropriate aliquot, adding to it 1 ml of
cone. HNO?, 2 ml 30% H202 and diluting to 100 ml with
deionized distilled water.
Nickel Nitrate
(1% Ni solution) — Dissolve 4.956 g of ACS reagent
grade Ni(N03)2-6H20 in 100 ml of deionized distilled
water.
Nickel Nitrate
(5% Ni Solution) — Dissolve 24.780 g of ACS reagent
grade Ni(N03)2-6 H20 in 100 ml of deionized distilled
water.
SAMPLE PREPARATION AND PROCEDURE
Detailed procedures on sample preparation and the final
concentration of nickel depend on sample type, matrix, and
concentration of selenium to be determined. In all cases
where total selenium is to be determined, the sample is
subjected to vigorous oxidation to solubilize the selenium.
Well and Surface Water
Transfer 100 ml of well-mixed sample to a 250-ml Griffin
beaker, add 3 ml cone, redistilled HNO-j and 5 ml 30%
H=02. Heat for one hr at 95 °C or until the volume is
slightly less than 50 ml. Cool and bring back to 50 ml with
deionized distilled water. Pipet 5 ml of this digested solu-
tion into a 10-ml volumetric flask, add 1 ml of the 1%
nickel nitrate solution and dilute to 10 ml with deionized
distilled water. The sample is now ready for analysis.
Since the nickel concentration is 0.1%, the sample
should be compared to a standard curve constructed from
standards also containing 0.1% nickel. The aliquot size
used for injection into the furnace should be the same for
both samples and standards. Recommended volume for
this type of sample is 25 to 100 ul.
To verify the absence of matrix or chemical interference,
an aliquot of the digest solution should be spiked with a
known amount of selenium, nickel nitrate added, and
analyzed in the same manner. The actual signal compared
to the expected response will indicate the presence of any
significant interference. Those samples where interference
is detected must either have the interference reduced by
dilution or be analyzed by the method of standard addi-
tions. (See discussion on interferences.)
Many surface water samples having low dissolved solids
(400 mg/'l) may be concentrated 5X during the digestion
step. Even though this technique extends the detection
'Since there are differences between individual furnaces and the
reading and setting of the maximum allowable charring tempera-
ture, each furnace should be checked to determine the maximum
charring temperature before beginning the analyses.
110
ATOMIC ABSORPTION NEWSLETTER
VoL 14, No. 5, September-October 1975
-------
limit, the solution must still be checked for interference by
spiking an aliquot of the concentrate and performing the
analysis.
Samples with sulfate concentration higher than 200
mg/1 should be analyzed in the presence of 1% Ni. Samples
are prepared and analyzed as previously described except
the addition of nickel nitrate should be 2.0 ml of the 5%
Ni solution diluted to 10 ml. Results should be determined
from a standard curve prepared from standards containing
1% Ni.
Industrial Waste Effluent
Sample should be prepared in the same manner as sur-
face water, but the nickel concentration in the final dilution
should be 1%. Results of many industrial effluents can be
determined from a standard curve, but again each must be
checked for possible interferences before this assumption
can be made. In some cases sample dilution may be re-
quired. A typical set of data from Se standards in a 1%
Ni matrix is listed in Table I. If it is necessary to use the
method of standard additions, the size of the aliquot used
for injection into the furnace will depend on the reproduc-
ibility of signal response and the amount of interference
encountered. An aliquot of 25 ul was used for the work on
this type of sample as reported in this paper.
Sediments and Sludges
Weigh and transfer to a 250-ml Griffin beaker a 0.5-g
portion of a sample which has been dried at 60°C, pul-
verized, and thoroughly mixed. Add 5 ml of cone. HNOj
and cover with a watch glass. The sample should be heated
at 95°C and refluxed to near dryness. Allow the sample to
cool, add another 5 ml of cone. HN03 and repeat the
digestion step. After the second reflux step has been com-
pleted, allow the sample to cooL add 3 ml of cone. HN03
and 10 ml 30% H202. Return the beaker to the hot plate
for wanning to start vigorous reaction. When the reaction
has commenced, immediately remove the beaker from hot
plate. After effervescence has subsided, return the covered
beaker to the hot plate and reflux for 15 minutes. After
cooling, dilute the sample to 50 ml with deionized distilled
water. Mix and withdraw a 5-ml aliquot, to be diluted to
10 mL for analysis by the method of standard additions.
Each final solution should contain 1% Ni (2.0 ml of 5%
Ni solution) and suspended solids should be permitted to
settle before analysis. It is suggested that solutions used
for analysis by the method of standard additions contain
5 ml of sample plus 15, 30, and 45 ug Se/1. Because the
possibility of encountering severe interferences is greatest
in this type of sample, a 5-ul aliquot should be used for
furnace injection. A detection limit of 5 }ig/g sample can
be achieved with this method using a 5-jjil injection.
ANAIYT1CAL PROCEDURE
The instrument should be operated using the conditions
as listed in the section on Equipment. As previously men-
tioned, the size of the aliquot used for furnace injection
will depend on the sample type as well as the matrix. When
the method of standard additions is required, a linear
curve over the entire range of the additions is necessary for
the results to be considered valid.
The life and performance of the furnace tube using this
method will mainly be affected by the number of analyses
completed. Many tubes have lasted for more than 100
firings but it is recommended, because of varying sample
types, that the tube be replaced after 100 firings. Prolonged
use of a given tube will result in elevated values, sometime
exceeding the expected value by more than 10%; but since
the increase is a gradual drift, there is no loss of precision
in consecutive analyses. For those samples which have a
complex matrix including metals of low volatility and re-
quiring the 1% nickel matrix, a conditioning cleaning burn
after each analysis may prove helpful. This can be accom-
plished by eliminating the drying and charring step, and
atomizing at 2700°C for 15 sec without gas interrupt
RESULTS AND DISCUSSION
Effect of Nickel
The addition of nickel to the sample serves three pur-
poses. First, it is believed to form a stable selenide com-
pound at the beginning of the char cycle thereby reducing
the volatility of selenium. This allows the use of elevated
charring temperature for complete ashing and volatiliza-
tion of some possible interfering and non-specific absorb-
ing substances. The second advantage of the nickel is the
increase in sensitivity gained because of the enhancement
effect. To demonstrate this effect a new graphite tube was
placed in the furnace, and the charring temperature was
set at 200°C. Average absorbances of 0.110 and 0.225 re-
spectively were recorded for a 25-ul injection of a 40 ug
Se/1 solution first without, and then with the addition of
nickel (1000 mg/1). The absorbance value (0.225) for the
charring temperature of 200° C is, for all practical pur-
poses, the same as when the charring temperature for the
same selenium solution is raised to 1500°C (Abs=0.235i.
This comparison, with and without the nickel, indicates
about a two-fold enhancement because of the nickel. This
enhancement may be due to a decrease in the rate cf
atomization or a change in the efficiency of the atomization.
Under the standard conditions given, a nickel concentration
of 100 mg/1 to 2000 mg/1 gives a similar enhancement
but when the nickel concentration is increased to 10,000
mg/1, or 1%, the absorbance for 25 ul of a 40 ug Se/1
solution drops to 0.170. Since the amount of nickel in the
furnace during atomization is critical to the signal re-
sponse, it must be controlled and the same quantity must
be present for both standards and samples.
Thirdly, the nickel serves as a stable matrix for those tap.
surface and well waters which have low concentrations of
trace metals and sulfate ion. thereby permitting the use
of a standard curve of the same nickel concentration.
Standard Curve
Table 1 shows the average absorbance and relative
standard deviation values for a composite standard curve
in a 0.1% Ni matrix over a concentration range of 5 to
100 [ig Se/1. The volume of the aliquot used for the injec-
tion for each standard was 25 ul. These standard data are
the result of values collected on 9 different days over a
period of 3 weeks. The composite data reflect normal dailv
variations in instrumental parameters and the effect of
different graphite tubes. Selenium is linear up to an ab-
sorbance value of 0.4 in a 0.1% nickel matrix. A working
detection limit using this technique is 2 (ig Se/1. This
detection limit can be extended to 1 ug Se/1 using a 100-ul
aliquot injection or to as little as 0.2 ug Se/1 if the sample
is first concentrated five times by evaporation, and a 50-ul
aliquot used for the injection. In both cases the concentra-
tion of the constituents in the sample matrix will be the
determining factor. To verify this procedure and standard
data, quality control check samples supplied by the Quality
Assurance Branch of the Environmental Monitoring and
ATOMIC ABSORPTION NEWSLETTER
Vol. 14, No. 5, September-October 1975
111
-------
TABLE I
Selenium Standard Data in 0.1 % Ni Matrix'
Se Concenrrafion
ug//iter
5
10
20
40
50
100
Average
Afasorbonce
0.035
0.065
0.118
0.235
0.290
0.540
% Relative
Std Deviation +
14.2
11.6
9.3
7.2
6.4
4.1
the 95% confidence level.
Selenium Standard Data in 1 % Ni Matrix*
Se Concentration
ug//iter
10
20
40
50
100
Average
Afaiorbance
0.046
0.091
0.170
0.219
0.413
*25-u,l sample aiiquots.
Support Laboratory in Cincinnati were analyzed at the 4,
16, and 48 ug/l levels with recoveries of 90%, 97%, and
96% respectively.
Water Matrix — Sulfate Interference
A major concern of any analytical technique is the
possible effect of the common minerals and anions present
and their concentration on the analytical result. Table II
lists a variety of parameters, their concentrations, and
the selenium response observed. Examination of the data
indicates an inverse relationship between the concentration
of magnesium and sulfate and the selenium absorbance.
Since it is known that a concentration of magnesium as
high as 200 mg/1 has no effect on the selenium response,
the increase in the suppression of selenium is attributed to
the increasing concentration of sulfate. Table HI is also
evidence of sulfate interference. Section (A) shows the
effect of large concentrations of sulfate. Section (B) shows
the effect in more detail over a small concentration range.
Section (C) shows that the degree of the sulfate suppres-
sion can be reduced by increasing the amount of nickel
present during the analysis. If the concentration of the
nickel in the injection aliquot is increased to 1% (10,000
mg/1), an injection of 50 ug of sulfate (50 ul of 1000 mg
SOt/l) will only cause a 15% suppression to the signal
generated by 1 nanogram of Se (25 ul of 40 |ig Se/1). See
Table V for a comparison.
Chloride and Nitrate Interference
Chloride and nitrate also affect selenium absorption. In
both the 0.1% and the 1% nickel matrix, chloride concen-
trations greater than 800 mg/1 cause a significant suppres-
sion (greater than 5%) of the absorbance. If the chloride
is increased to 2000 mg/1, the suppression in 0.1% Ni and
1% Ni is approximately 15% and 30%, respectively. Thus
an increase in nickel concentration does not decrease the
suppressive effect of chloride, and therefore the method as
described is not applicable to the analysis of seawater
and brines.
In selenium solutions containing 1% v/v cone. HNQs
there is an 80% reduction in the Se absorbance when the
nickel is omitted, but in a 0.1% Ni matrix with. 3% v/v
cone. HN03 no reduction was observed. At levels above 3%
nitrate, interference is encountered in the 0.1% nickel
TABLE II
Selenium Absorbance in Six Synthetic Surface Water Matrix Solutions
of Various Concentrations*
Element, Anion
or Measured
Parameter
Calcium
Magnesium
Sodium
Potassium
Alkalinity
Chloride
Total hardness
Total dissolved solids
Sulfats
Volume of Aliquot
25 ul
% response
50 ul
% response
lOOul
% response
Distilled
Water •
Solution
0
0
0
0
0
0
0
0
0
0.122
100%
0.224
100%
0.345
100%
Concentration mg/liter
I
90
21
82
16
180
174
310
570
84
—
—
0.346
100%
2
180
25
210
32
280
350
550
1200
260
Se
—
0.223
100%
0.321
93%
3
180
41
260
32
280
350
630
1450
440
4
180
41
390
32
560
350
600
1760
440
5
360
41
390
63
560
700
1200
2300
440
6
360
82
520
63
560
700
1240
2900
870
Absorbanca Values
0.104
85%
0.181
81%
0.269
78%
0.100
82%
0.189
84%
0.263
76%
0.109
89%
0.191
85%
0.279
81%
0.097
80%
0.164
73%
0.237
69%
'Each of the 6 synthetic matrix solutions and the distilled water solution contained 20
u.g Se/1 in 0.1% Ni. 1% v/v cone. HNO*. 2^0 v/v 30% H30S.
112
ATOMIC ABSORPTION NEWSLETTER
Voi. 14, No. 5, September-October 1975
-------
TABU III
Effect of Suifata on Selenium Absorbance
Volume of
Suifate
Aliquot
0
25 |il
(A) 25 |il
25 ul
25 ul
0
25 ul
50 uJ
25 ul
(B) 50 ul
100 |il
50 ul
100 |il
0 (it
25 |il
50 jil
(C) 50 uJ
50 ul
75 ul
Concentration
0
500
1000
2000
4500
0
270
180
450
270
180
450
270
0
450
270
360
450
360
Total ug
SO4//n/ection
0
13
25
50
113
0
6.8
9.0
11.3
13.5
18.0
22.5
27.0
0
11.3
13.5
18.0
22.5
27.0
Se
Absorbance*
0.30
0.22
0.16
0.10
0.07
Se
Absorbancef
0.230
0.230
0.215
0.199
0.166
0.140
0.126
0.107
Se
Absorbance +
0.230
0.212
0.201
0.187
0.168
0.142
% Suppress/on of
Se Absorbance
—
27%
45%
66%
76%
—
0
6%
13%
28%
39%
45%
53%
—
8%
13%
19%
27%
38%
*Se absorbance value and corresponding suppression is the result of a 25-ul injection of SO ng Se/1
in 0.1% Ni (25 ug iYi/irt;ec«on;, 1% v/v cone. HNOj 2% v/v 30% HiO» with the listed quantity
of suifate pipetted on top of the Se injection.
tSame conditions as in * except 25 ul of 40 |ig Se/1 in 0.1% Ni f2$ pg Ni/injection), 1% v/v
cone. HNOs. 2% v/v 30% H»05 was used for injection.
"Same conditions as in * except 50 (il of 20 |ig Se/1 in 0.1% Ni (SO \ig Hi/injection), 1% v/v
cone. HN03,2% v/v 30% H»0j was used for injection.
matrix. Although this interference can be somewhat re-
duced and stabilized by the use of a longer charring cycle,
concentrations of over 30,000 mg NOa/l should be avoided.
Single Metal Interference
Table IV lists concentrations of single metal solutions
and the degree to which these metals affect the Se absorb-
ance in 0.1% Ni matrix. These approximate results are
given as an indication of when the analyst can no longer
reliably use a standard curve prepared in 0.1% Ni for the
determination. Special attention should be given to the
concentration of Fe, Sn, Si, Al, Mn, V, and Cr. Although
seldom present at these concentrations in surface and tap
water, there may be other types of environmental samples
including sludges and sediments where these elements may
exist in even greater concentrations than listed. It has been
determined that increasing the concentration of the nickel
to 1% decreases the suppressive effect of many metals. A
comparison of the suppressive effect of some of the more
critical metals in 0.1%.and \% nickel matrix is given in
Table V. Although the selenium response is lower in 1% Ni
than in 0.1 ^ Ni when other metals are absent, the same is
not necessarily true with the addition of these metals ,is evi-
dent in Table V. This phenomenon can be an advantage in
eliminating large suppressive effects when analyzing sam-
ples with a complex matrix.
Since all of the metals tested have a concentration which
can be tolerated without causing an interference, the deter-
mination of their composite effect at those concentrations
both with and without the synthetic water matrix was
important. Table VI lists the matrix parameters and trace
metals, their concentrations, the affected selenium absorb-
ance, and percent suppression in both 0,1% and 1% nickel
solutions. In reviewing Table VI, it is apparent that there
is a composite effect in 0.1% Ni but not in 1% Ni and that
the combination of matrix and trace metals produces in-
creased suppression in 0.1% Ni. This suppressive effect is
strong evidence for using 1% nickel when analyzing sam-
ples with a complex matrix or ones that contain ions or
trace metals at concentrations known to interfere. The
sample may be analyzed using either a standard curve
prepared in 1% Ni or, if necessary, by the technique of the
method of standard additions. Whenever possible, and
especially for tap water and clean, low dissolved solids
surface water, the 0.1% nickel matrix should be used be-
cause of the added sensitivity.
Dissolved and Suspended So/ids
In considering the effect of dissolved and suspended
solids, it was determined that the nature or chemical com-
position of the solids rather than the physical state was the
important factor. Also the amount of an interfering sub-
ATOMIC ABSORPTION NEWSLETTER
Vol. 14, No. 5, September-October 1975
113
-------
TABLE IV
Suppression Effects on
Element
Ag
Al
As
3
Bo
Be
Cd
Co
Cr
Cu
Fe
U
Mo
Mn
Ni
P
Pb
Sb
Si
Sn
Sr
Ti
Tl
V
In
Concentration
W/TIC/I HOJ
No Effect,
mg/l
400
20
40
300
400
10
400
400
50
100
4
300
200
20
300
100
200
40
10
2
400
200
40
20
400
Selenium Absorbance* of Single Metal Solutions +
Concentration
mg/l
_
40
100
400
—
20
—
—
100
200
10
400
400
50
400
200
300
50
20
4
—
400
200
30
—
Suppress/on
10%
10%
10%
5%
10%
10%
10%
10%
10%
10%
10%
25%
15%
10%
10%
10%
5%
20%
10%
Concentration
mg/l
—
200
400
—
—
40
—
—
200
—
20
—
—
200
—
400
400
100
40
10
—
—
400
100
—
Suppress/on
65%
50%
20%
55%
30%
50%
75%
30%
20%
50%
40%
30%
30%
•A 25-M.l injection of 40 n« Se/1 in 0.1% Ni, 1% v/v cone. HNOj, 2% v/v 30% H,0» was used for this
comparative work.
'The approximate suppression values are the result of a 25-nl injection of the listed concentrations
being pipetted on top of the selenium injection.
TABLE V
Comparison of Suppression Effects of Trace Metal Solutions
on Selenium Absorbance in 0.1% Ni and 1% Ni Solutions
Trace Metal Solution
25u/40ugSe//
;n0.7%N/
50 ul
50 ul
50 ul
50 ul
50 ul
50 ul
50ul
50 ul
_
lOOmgAl/l
100mgCr/l
1QOmgCu/l
lOOmgFe/1
lOOmgSi/l
100 mg Sn/l
100mgV/l
1000mgSO4/l
Absorbance
0.235
0.075
0.095
0.195
0.120
0.025
0.015
0.085
0.047
% Suppression
—
68
60
17
49
89
94
64
80
Absorbance
0.170
0.105
0.140
0.154
0.126
0.086
0.103
0.148
0.145
% Suppression
_
38
18
9
26
49
39
13
15
ATOMIC ABSORPTION NEWSLETTER
Vol. 14, No. 5, September-October 1975
-------
TABLE VI
Comparison of Selenium Absorbance in 0.1 % Ni and 1 % Ni Solutions
Containing Synthetic Matrix and Trace Metals
Matrix
TVaco Metals
Solution
Content
Se
Se + matrix
Se + metals
Se + matrix
•4* metals
Element
Calcium
Magnesium
Sodium
Potassium
Alkalinity
Chloride
Total hardness
Total dissolved
solids
Sulfate
Aliquot
Injection
Volume
25 ul
25 uJ
25 uJ
25 uJ
Cone.
180 mg/l
20 mg/l
190 mg/l
32 mg/l
280 mg/l
350 mg/l
600 mg/l
1160 mg/l
220 mg/l
0.7%
Element
Sn
Be
Fe
Si
V
Mo
Al
As
Cu
Mn
P
N;
Se C30 \ig/l) % Enhanc. or
Absorbance Suppression
0.170
0.175
0.150
0.106
—
+ 3%
-12%
-38%
Cone.
1 mg/l
2 mg/l
2 mg/l
5 mg/l
5 mg/l
10 mg/l
20 mg/l
20 mg/l
20 mg/l
20 mg/l
20 mg/l
7%Ni
Se (24 ug/U % En/ianc. or
Absorbanca Suppression
0.105 -
0.101 -4%
0.111 +6%
0.109 +4%
stance present during atomization is the important consid-
eration — not whether injected as a dissolved or suspended
solid. Review of Table II reveals that a change in total
dissolved solids in the synthetic water matrix from 1450
mg/l to 2300 mg/l did not produce a significant difference
in the selenium absorbance. Since the total dissolved
solids for most surface water are below 2000 mg/l, solids
should not be a problem in the analysis of water samples
provided that the non-specific absorption does not exceed
the background correction capability of the instrument.
Sample Analysis and Recovery Data
Analytical results and spike recovery on a variety of
sample types are given in Table VII. The sample prepara-
tion used was that described in this paper. The results were
determined from standard curves prepared in 0.1% and
1% nickel solutions, and by utilizing the method of stan-
dard additions.
CONCLUSION
This method utilizing the HGA-2000 Graphite Furnace
provides a rapid procedure for analyzing a variety of water
samples for selenium. After the sample preparation and
solubilization. the samples are diluted in either 0.1% or
1% nickel matrix and compared to a standard curve of the
same matri* to determine the result of the analysis. The
method demonstrates satisfactory precision and is suffi-
ciently sensitive with a working detection limit of 2 ug
Se 1 which can be extend
-------
TABLE VII
Results of Sample Analyses and Recovery Data for Selenium
No. of
Sample Type Somp/es
Anoryzed
Tap water
Surface water
Well water
(N. Mexico)
Well water
(N. Mexico)
Drinking water
for animal ex-
posure studies
Sewage plant
effluent
Industrial
waste effluent
Landfill
leachate
Ocean dis-
posal
Sludges
Sediments
Solid gelatin
ref. std.
2
2
6
4
4
2
6
3
1
1
1
2
3
1
No. of
PosmVe
Occur-
rences
0
0
1
0
4
0
0
0
0
1
0
0
0
1
Technique
Amount of
Detected Deter-
mination
N.D. < 2 ug/l
N.D. <2 ug/l
5ug/I
N.D. <0-5 ug/l
57 ug/l
58 ug/l
o?M/!
N.D. <2ug/I
N.D. < 5 ug/l
N.D. < 4 ug/l
N.D. < 10 ug/l
50 ug/l
N.D.
------- |