x°/EPA
United States
Environmental Protection
Agency
Environmental Monitoring and Su
Laboratory ~*
Cincinnati OH 45268
Research and Development
EPA-600/S4-81-043 Aug. 1981
Project Summary
An Evaluation of the
Hewlett-Packard 5993B Gas
Chromatograph/Mass
Spectrometer
Denis L Foerst
This report gives the results of an
evaluation of a Hewlett-Packard Model
5993B computerized gas chromatog-
raph/mass spectrometer. The evalua-
tion was done according to the pro-
cedures found in EPA research report
number EPA-600/4-80-025. "Per-
formance Tests for the Evaluation of
Computerized Gas Chromatography/
Mass Spectrometry Equipment and
Laboratories." Changes in the "Per-
formance Test" specifications for 4-
bromofluorobenzene (BFB) are noted
in the report.
The system performed satisfactorily
during the evaluation and is acceptable
for the analyses of organic compounds
using methods similar to EPA Methods
624 and 625. The HP-5993B system
produced valid mass spectra, exhibited
good to excellent stability, and gave a
satisfactory detection limit. The pre-
cision and accuracy for external or
internal quantitation using direct sam-
ple injection or purge and trap were
excellent. Both 14 AMU and PBS
mass spectral library searches were
accurate; however, the search times
using the PBS library search were
longer than the requirements specified
in the "Performance Tests." Some
problems, particularly with brominated
compounds, were discovered during
this evaluation. The manufacturer
corrected the instrument to alleviate
these problems.
This Project Summary was devel-
oped by EPA's Environmental Moni-
toring and Support Laboratory, Cin-
cinnati, OH, to announce key findings
of the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The quality of environmental mea-
surements performed using gas chro-
matography/mass Spectrometry (GC/
MS) can be affected by the performance
of the GC/MS system. EPA recently
published a series of ten performance
tests which can be used to describe and
evaluate the performance of GC/MS
systems. A system that achieves the
minimum performance criteria specified
in these performance tests can be
considered capable of producing reliable
quantitative and qualitative data.
These performance tests were con-
ducted on an HP-5993B GC/MS system
to determine if this relatively low cost
GC/MS equipment is practical for use
in performing environmental analyses.
Procedure
Eight of the ten tests were carried out as
described in the "Performance Tests."
Test seven was not performed because
the test is not applicable to equipment
only and test ten is applicable only for
GC/MS systems which have a solid
-------
probe. The HP-5993B system evaluated
contains: an HP-5992B GC/MS core,
with an electron impact ionization
source; an HP 21MXE Computer, with
32K resident core operated under SIDS
software; an HP-7900A Disc Drive, with
five megabytes total storage; an HP-
2648A Graphics terminal; an HP-7675A
Purge and Trap Unit; a glass lined jet
separator; and a Tektronics 4532 Video
Hard Copy unit. This system is also
equipped with an optional magnetic
tape drive and optional incremental
plotter. During the evaluation, the
system experienced difficulties in pro-
ducing bromine isotope ratios that were
within ± 3% of the theoretical value.
The manufacturer corrected the problem
satisfactorily.
Results
After the bromine isotope ratio problem
was corrected, the system achieved all
the minimum criteria given in the "Per-
formance Tests." The following sum-
marizes the description of each test, the
minimum performance specified and
the results obtained on the HP-5993B.
Test 1 - Spectrum Validation. This test
evaluates the ability of the system to
produce a mass spectrum which achieves
all the key ion abundances for both
decafluorotriphenyl phosphine (DFTPP)
and 4-bromofluorobenzene (BFB). The
key ion abundance criteria are given in
Table 1. The criteria for BFB have been
changed since publication of the per-
formance tests. The average abundance
of each key ion as summarized in Table 1
indicates that the system does achieve
acceptable spectrum validation. Use of
the software tuning program CNORM is
necessary to achieve the relative abun-
dance criteria for DFTPP but is not ne-
cessary for the BFB relative abundance
criterion. The resolution criteria were
always achieved before the program
CNORM was used.
Test 2 - System Stability. The GC/MS
system must be stable enough to give
acceptable spectrum validation tests for
a period of at least 24 hours while
maintaining identical tuning para-
meters. The HP-5993B system gave
acceptable DFTPP spectrum validation
tests after a 20 hour period and a 64
hour period. The system gave acceptable
BFB spectrum validation tests after a 40
hour period. During the 40 hour period,
16 half-hour runs, totaling 9800 repet-
itive scans were collected demonstrating
that the HP-5993B system is very stable
during use.
Test 3 - Instrument Detection Limit.
The instrument detection limit is com-
puted from the relative abundance of
certain DFTPP key ions divided by the
relative abundance of contaminant
ions. The minimum criteria is an in-
strument detection limit of 50 ng of
DFTPP for systems used for the analysis
of industrial or municipal waters and
wastes and 30 ng of DFTPP for systems
used for the analysis of ambient or
drinking waters. The HP-5993B gave an
instrument detection limit of 25 ng of
DFTPP; thus the system meets the
requirements for the analysis of indus-
trial and municipal wastewatersandthe
requirements for the ana lysis of ambient
and drinking waters.
Test 4 - Saturation Recovery. This test
evaluates the capability of the GC/MS
system to give good qualitative mass
spectra of small peaks eluting on the tail
of very large peaks. An aliquot of 11 //g
of 4-bromobiphenyl and 44 ng of DFTPP
was analyzed. The DFTPP eluted 1.2
minutes after the 4-bromobiphenyl and
gave an acceptable spectrum validation
test. There were no ions at m/z 76,152,
232, or 234, the most abundant ions of
4-bromobiphenyl in the mass spectrum
of DFTPP. The test requires that the
relative abundance of any 4-bromobi-
phenyl ions in the DFTPP mass spectrum
be below 5%.
Test 5 - Precision Test. The relative
standard deviation (BSD) for quantitative
analysis using repetitive acquisition of
mass spectra must be less than 7% to
meet the minimum criteria for this
Precision Test. The RSD reported must
be the average RSD from the analyses of
a solution of at least seven analytes
each at 20/ug/mL The analytes selected
must include at least one chlorinated
hydrocarbon that could decompose on a
hot metallic surface and one polynuclear
aromatic hydrocarbon of molecular
weight greater than 200. Two solutions
were prepared. The first contained
1,2,4-trichlorobenzene, hexadecane,
dimethylphthalate, hexachlorobenzene,
aldrin, dibutylphthalate, and pyrene. It
was found that the use of CNORM
hardly affects precision, because the
RSD after eight replicate analyses was
2.35% before treatment by CNORM and
was 3.43% after treatment by CNORM.
The RSD for the automated integration
of the same eight replicates was 4.26%
using external standard quantitation
and 3.06% using internal standard
quantitation.
A second precision test was performed
using direct injections of a solution ^
containing eleven analytes amenable to m
purge and trap. Chloroform, carbon
tetrachloride, bromodichloromethane,
trichloroethene, benzene, bromoform,
Table 1. DFTPP and BFB Spectrum Validation Test Results
Compound
DFTPP
BFB
Ion
51
127
198
275
365
442
68/69
70/69
197/198
199/198
441/443
443/442
50
75
95
174
96/95
173/174
175/174
177/176
176/174
Criterion
30% to 60%
4O% to 60%
1 0OP/o
10% to 30%
> 1%
>40%
< 2%
< 2%
< 1%
5% to 9%
> Oand<100%
17% to 23%
15% to 40%
30% to 60%
100%
>50%
5% to 9%
< 2%
5% to 9%
5% to 9%
95% to 101%
HP-5993B
Results'
43.6%±3.88
51.O%±3.38
100%±O.OO
20.3%±1.13
1.58%±0.39
62.1%±12.0
0.59%±0.72
0.27%±0.61
0.20%±0.30
5.76%±0.27
64.2%±3.66
18.7%±1.13
19.8%±2.38
40.8%±2.96
100%±0.00
84.6%±12.2
6.81%±0.74
0.79%±0.86
7.97%±0.52
6.76%±0.93
99.7%±0.91
"Average of 14 runs for DFTPP and 8 runs for BFB.
4
-------
t
1,1,2,2-tetrachloroethane, toluene,
chlorobenzene, ethylbenzene, and 4-
bromofluorobenzene were used. This
precision test gave an average RSD for
automated integration of 3.36% using
external standard quantitation and
2.91 % using internal standard quantita-
tion.
Test 6 - Library Search Test. The
"Performance Tests" require a mean
search time of 60 seconds or less when
a library of 10,000 or more mass spectra
are searched. Search time includes the
time for background correction but not
the time for listing the hard copy of the
results. Either the compound searched,
or an isomer, must be identified as most
probable.
The solvent extractable compound in
Test 5 plus 1,2-dichlorobenzene and
benzo(ghi)perylene were used for this
test. The background correction averaged
55 seconds when done manually and 6
seconds when done by automated soft-
ware. Two library search algorithms
were tested, the 14 AMU Search and
the Probability Based Search (PBS). The
mean search time, including automated
background corrections, was 53 seconds
using the 14 AMU software and 118
seconds using the PBS software. Both
sets of software gave the compound, an
isomer, or a compound of similar mass
spectra ranked first.
Test 8 - Quantitative Analysis and
Inert Gas Purge and Trap. This test
evaluates the precision for purge and
trap analysis, as was done in Test 5, and
the purging efficiency characteristic of
the purge and trap hardware interfaced
to the GC/MS. The minimum require-
ments are that the average purging
efficiency for 100 ng of each analyte
studied be greater than 70% with no
individual value less than 30%, the
average accuracy be between 90% to
110%, and that the average RSD be less
than 19%. The analytes used must
include chloroform, bromoform, 1,1,2,2-
tetrachloroethane, and 4-bromofluoro-
benzene.
The purge and trap analytes used in
Test 5 were also used in this test. The
average purging efficiency was 87%
with a range to 58% to 96%. The accuracy
and precision were 100.3% with an RSD
of 6.54% using external standard quan-
titation and 100.3% with an RSD of
3.90% using external standard quanti-
tation.
Test 9 - Qualitative Analysis with Real
Samples. This test evaluates performance
using a real environmental sample.
Since matrix effects can be very unpre-
dictable, no accuracy and precision for
quantitation is evaluated, only the
accuracy for qualitative analysis is
evaluated. The minimum requirements
are that the analytes dosed into the real
sample be qualitatively identified as the
actual analyte or as an isomer with a
similar 70 volt mass spectrum.
The final extract obtained from an
NBS sediment was used in this test.
Analysis was performed before and
after dosing the extract with acenaph-
thylene, fluorene, pyrene, chrysene,
benzo(a)pyrene, and benzo(ghi)perylene.
The PBS library software was used and
neither acenaphthylene nor fluorene
were found in the extract, but both were
correctly identified in the dosed extract.
The other four polynuclear aromatic
hydrocarbons were correctly identified
in both the extract and the dosed extract.
Conclusions
The HP-5993B GC/MS system meets
the minimum criteria for GC/MS systems
established by the EPA for the analysis
of environmental samples. In many
cases, the performance of the GC/MS
system can be rated excellent when
compared to these minimum criteria.
This type of low cost GC/MS system can
be routinely used with EPA methods
similar to Methods 624 and 625.
The EPA author is Denis L. Foerst of the Environmental Monitoring and Support
Laboratory, Cincinnati, OH 45268.
W. L. Budde is the EPA Project Officer (see below).
The complete report, entitled "An Evaluation of the Hewlett-Packard 5993B Gas
Chromatograph/Mass Spectrometer,"(Order No. PB 81-218 315; Cost: $8.OO.
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
A U.S GOVERNMENT PRINTING OFFICE 1981 -757-012/7266
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
Pi> 0000
U i> fiivviK FKUTtCTIUN
KtblU\! 5 LidK4KY
230 & DEAHeUKN STREET
CHICAGO IL 60604
------- |