United States
                             Environmental Protection
                             Agency
  Office of
  Solid Waste and
  Emergency Response
Publication 9380.5-14FS
PB94-963509
EPA 540/F94/056
September 1994
                             Indoor Air  Analysis  of  Volatile
                             Organic Compounds
                             Using Capillary Column
                             Gas  Chromatography/Mass Spectrometry
Office of Emergency and Remedial Response
Emergency Response Division
Environmental Response Branch MS-101
                          Quick Reference Fact Sheet
Introduction
The quality of indoor air and the resulting health risk asso-
ciated with  some potential exposure to volatile organic
compounds  (VOCs) from indoor air have become major
concerns to  building occupants, especially office workers.
To effectively address these concerns, indoor air monitor-
ing programs are needed  which  can produce data  of
known quality. The ensuing analytical method developed
by   the  United  States   Environmental  Protection
Agency/Environmental Response Team (U.S. EPA/ERT),
in conjunction with the Response Engineering and Ana-
lytical Contract  (REAC),  provides rapid turnaround  of
sample results while incorporating rigorous quality assur-
ance/quality control (QA/QC) procedures  to ensure reli-
ability of the analytical data.
The  gas  chromatography/mass  spectrometry  (GC/MS)
method was developed to provide support for the U.S.
EPA/ERT indoor air quality program. The GC/MS method
was designed to identify and quantify  45 target  VOCs
(i.e., aliphatic, aromatic, and chlorinated hydrocarbons) as
well as identify non-target VOCs related to indoor air en-
vironments.

The Analytical Method
Multi-residue trace organic analyses  for air samples are
greatly dependent upon MS for unambiguous identifica-
tion and confirmation of volatile and semi-volatile com-
pounds. It is the preferred detection method because  of
limitations in selectivity of alternative micro-analytical
techniques.  The proposed indoor air method uses a full
scan GC/MS and a fused silica capillary column for ana-
lyzing ambient air samples collected on charcoal tubes.
The method was evaluated on charcoal tubes spiked with
known quantities of 45 target VOCs and indoor air sam-
ples collected from office and home environments. It was
also evaluated by determining desorption efficiencies, re-
covery studies, linearity range, method detection limits
(MDLs), accuracy, and reproducibility. Identification and
quantification of compounds tested were made by compar-
ing  retention times and mass spectral data of unknown
compounds with  known compounds  from  calibration
standards.

Materials - Stock standard solutions of 45 target VOCs
were prepared by  accurately weighing 500 ±0.1 mg of
each neat compound into a 25 mL volumetric flask and di-
luting to mark with benzene free carbon disulfide (CSa).
Working and daily standards were prepared by serial dilu-
tion of the 20 mg/mL standard mixture with CS2. Standard
stock solutions of decafluorotriphenylphosphine (DFTPP)
and isotopically labeled (cyclohexane-di2, toluene-ds, 1,4-
dichlorobenzene-d4, and naphthalene-ds) internal stand-
ards (IS) were prepared  at  10,000  |J.g/mL each  by
weighing 100 ± 0.1 mg of DFTPP and IS into two separate
10 mL volumetric flasks and diluting to mark with dichlo-
romethane  and  CSa,  respectively. Secondary  solution
standards at 1000 ng/mL were prepared from the stock so-
lution by serial dilutions. The stock and secondary solu-
tions were stored  at -10°C to  -20°C in  Teflon®-lined
crimp-top amber bottles.

Sample Collection - Air samples from home and office
areas were collected on 600-mg two-stage  charcoal tubes
(400-mg front- and 200-mg back-activated  charcoal sepa-
rated with a foam plug) with Gillian monitoring pumps. A
flow rate of 2 L/min (for a total of 1,200 L samples) was
used throughout the study in order to achieve lower detec-
tion limits. The charcoal tubes were desorbed with 2 mL
CSa spiking 1 mL of the sample extract with IS and ana-
lyzed by a capillary column GC/MS.

GC/MS System - The analytical system consisted of a
HP5995A GC/MS equipped with an HP7673A autosam-
pler and controlled by an HP-1000 RTE-6/VM computer
data system equipped with the National Institute of Stand-
ards and Technology PBM Mass Spectral Library. A 30 m
x 0.32 mm ID, Rtx-5 (Restek Corp.), fused  silica capillary
column with a 0.50 |J.m film thickness was used to sepa-
rate the compounds tested. The analytical design was a to-
tally automated system for sample injection, analysis, data
reduction, and reporting.

-------
Method with Acceptance Criteria
The QA/QC procedures are integral to analytical methods
which produce data of  known  quality.  The  following
guidelines were used prior to initiating on-going data col-
lection, i.e., blanks, calibrating standards, or samples:
Tuning and Calibration - The GC/MS system was tuned
daily  or every 12 hours with 50 ng DFTPP. The results
were then checked to verify that the U.S. EPA tuning and
mass spectral ion abundance criteria were met. The system
was initially calibrated with a standard VOC mixture in
the concentration  range of 1 jig/mL to 100 ug/mL (1,5,
10, 25, 50, and  100 ug/mL)  using the internal standard
method; and was routinely checked by analyzing a con-
tinuing calibration standard (VOC mixture at 25 |ig/mL
with IS at 50 (ig/mL) to ensure that the initial instrument
sensitivity and linearity did not change significantly.

Acceptance Criteria - The data from the initial and con-
tinuing calibrations were used to calculate average relative
response factors (RRFj)  and continuing relative response
factors (RRFC), respectively. For the calibration  to  be
valid, a criteria of maximum percent relative standard de-
viation (%RSD < 30) and  a  minimum RRFj > 0.2 was
adopted for all target compounds  in the initial calibration.
For the continuing calibration check, measured RRFC must
be within ± 20% (i.e., %D < 20) of the mean values calcu-
lated during initial calibration. In  addition, all target com-
pounds must have a minimum RRFC > 0.2.
Desorption Efficiency (BE) - Desorption efficiency cor-
rections were applied for each compound detected from
samples  analyzed. The  DEs were  determined for  the
method Target Compound List (TCL) at  2,  10, 50, and
200 [Ag levels and were  established for every new lot of
carbon for each compound before  samples were analyzed.
Detection Limits, Accuracy, and Precision - The MDLs
for the compounds  tested were determined to  be better
than 0.2 parts-per-billion-by-volume (ppbv) at the 99 per-
cent confidence level. The linearity of the GC/MS system
(1 to 100 |ig/uL) was demonstrated by the initial six point
calibration data for  the target compounds. The accuracy
and reproducibility (precision) of the method, determined
by analyzing a set of four replicate  spikes at 50 ug level,
was found to be better than 80 percent and 10 percent, re-
spectively.

Applications
The indoor air method has been used for over five years to
analyze ambient air samples from indoor (homes and of-
fices) and outdoor (tire  fires) environments. The results
from an indoor air study in a northern New Jersey urban
office and  site  were compared with a VOC-INDO data
base [a national data base comprised of 320 VOCs (with
66 VOCs measured indoors and 321 measured outdoors)
from a distribution of volatile organic chemicals in indoor
and outdoor air]. Nearly 90 percent of the indoor air data
was from California and New Jersey. The outdoor air data
was  predominately  from  California, New Jersey, and
Texas.

The range of results for individual compounds varied from
not detected [ = 0.1 ppbv]  to less than 1.0 ppbv
(indoor air) and 36.0 ppbv (outdoor air).  The compounds
normally detected are either in the background outdoor/ur-
ban  air  (1,1,1-trichloroethane,   carbon  tetrachloride,
trichloroethene,  etc.) or from automobile exhaust (ben-
zene, toluene, and xylene).
The MDLs for the compounds tested were determined to
be better than 0.2 ppbv at the 99 percent confidence level.
The accuracy and reproducibility (precision) were better
than 80 percent and 10 percent, respectively. The QA/QC
procedures were found to be essential criteria for obtaining
reliable analytical data. Based on over five years of experi-
ence using the proposed indoor air method with full scan
GC/MS and fused silica capillary column,  the following
conclusions have been reached:

°   The analytical results obtained are usually close to the
    method detection limit values.

°   The method provided  quality  VOC data for indoor
    and outdoor air environments with quick turnaround
    (one day) for emergency response samples.

o   The results indicate that 1, 1, 2, 2-tetrachloroethane
    may undergo  in-situ degradation or reaction on char-
    coal tubes.

o   The advantage of GC/MS over GC/FID detection is
    that a  broad spectrum of target compounds can be un-
    ambiguously identified and quantified and  non-target
    compounds identified as tentatively identified com-
    pounds (TICs).
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.

            For more information, contact:
          Raj Singhvi, Phone (908) 321-6761
           (908) 321-6660 (24-hour hotline)

-------