Sampling for semivolatile organic contaminants in

environmental compartments

Lisa Rodenburg

The Universe of Non-polar Chemicals

PBTs

VOCs

POPs

VOCs = Volatile organic chemicals

SOCs = semivolatile organic chemicals

POPs = Persistent Organic Pollutants

PBTs = Persistent, Bioaccumulative, & Toxic

SOCs

Many classes of contaminants can be sampled and measured together:

PCBs = polychlorinated biphenyls = myriad uses, banned since 1970’s = PBTs

PAHs = polycyclic aromatic hydrocarbons = combustion by-products = less P, very BT

PBDEs = polybrominated diphenyl ethers = current use flame retardants = not very P, very B, not sure how T

OCPs = organochlorine pesticides (DDT, etc) = some in current use (α- and γ-HCH, α- and β-endosulfan ), some banned (DDT) = most are PBTs

PCBs

PCBs were previously sold as “Aroclors” and used as fluids in electrical equipment, particularly transformers and capacitors.

PCBs are classified as probable human carcinogens and have been shown to cause a range of serious non-cancer health effects in animals.

The manufacture, processing, and distribution in commerce of PCBs were banned in 1976 due to concerns over their toxicity and persistence in the environment.

About 1.3 million metric tons of PCBs were produced world-wide.

PCBs consist of 209 congeners, which may have 1 to 10 chlorines.

A group of congeners having the same number of chlorines is a “homolog group”

PAHs

PAHs are Polycyclic Aromatic Hydrocarbons

They contain 2 or more fused aromatic rings

Products of combustion of any type as well as evaporative emissions from fuels.

A small amount of PAHs are produced naturally (volcanoes, forest fires, etc.)

Humans are exposed to PAHs by breathing contaminated air (including tobacco smoke) and eating grilled foods.

The Department of Health and Human Services (DHHS) has determined that some PAHs may reasonably be expected to be carcinogens.

PBDEs

OCPs

Can have 1-10 bromines, numbered in same was as PCBs

Many different structures and numbers of chlorines DDT

heptachlor

2,2’,4,4’,5-pentabromodiphenyl ether = BDE 99

Outline

• Sampling– Air– Water– Other

• Cleanup– The easy way– The hard way

• Detection– GC/ECD– GC/MS

• Cautionary Tales

Sampling

Extraction

Cleanup

Detection

QA/QC

What are semivolatile contaminants?

• On the basis of vapor pressure, we can divide the nonpolar or slightly polar compounds into VOCs and SOCs.– VOCs = vapor pressure > 10-3 atm

not on atmospheric particles– SOCs = vapor pressure < 10-6 atm

significant fraction on atmospheric particles– Some things fall through the cracks, like

naphthalene.

VOCs (TO-15)Fluorinated, chlorinated, brominated C1, C2, C3 compounds

Mono- and di- chlorobenzenes

Methyl and ethyl benzenes

1,3-ButadieneAcetoneAcetonitrileAcetyleneAcrylonitrileBenzene  

Ethyl AcrylateEthyl Tert Butyl EtherEthylbenzeneHexachloro-1,3-ButadieneMethyl Ethyl KetoneMethyl Isobutyl KetoneMethyl MethacrylateMethyl Tert-Butyl Ethern-OctanePropyleneStyreneTert-Amyl Methyl Ether

 SUMA Canister sampling

SOCsPCBs

PAHs

Organochlorine pesticides

PBDEs

Others?

Sampled using a high-volume air sampler

The Hi-Vol

Filter Canister (containing PUF or XAD-2 sorbent)

Vacuum pump

Timer

Measure pressure drop before and after every sample. Calibration curve converts pressure drop to flow rate. Timer gives time, hence volume of air sampled.

Pitfalls• Breakthrough of more

volatile contaminants (minimize flow rate)

• Gas/particle partitioning(minimize flow rate)

• Detection limits(maximize flow rate)

• Motor instability(pre- and post-calibration)

• Contamination from motor, O-rings, etc.(keep everything clean, vent motor)

Measure breakthrough by occasionally cutting a PUF in half and analyzing the top and bottom separately.

Sorbent choice

• PUF allows greater breakthrough of polar and volatile compounds.

• XAD-2 has a huge PAH background, especially low MW PAHs.

• PUF can be very clean.

Run lots of blanks!

Breakthrough of PCBs on PUFPCB congener Percent Breakthrough

8,5 29 %

18 22 %

17+15 27 %

16+32 14 %

31 4 %

28 7 %

21+33+53 2 %

22 5 %

45 28 %

46 9 %

52+43 0 %

49 8 %47+48 1 %

44 1 %74 1 %

87+81 8 %177 3 %

202+171+156 100 %201 5 %

203+196 5 %195+208 15 %

194 3 %

Heavier PCBs, less breakthrough.

Breakthrough significant for PCBs with 3 or less chlorines

Blank contamination

PCBs can have 1 – 10 chlorines. PCBs are numbered such that higher numbers have more chlorines.

Water Sampling• Whole water or grab samples

– Detection limits require very large samples– Blank contamination a big problem– Volatilization

• Dissolved vs. Particulate– Filter for particles, sorbent for dissolved– Choice of sorbent is tough

• XAD-2 PAH contamination• Tenax, C18 cleanup problems

– Choice of platforms:• Infiltrex – expensive, unreliable• TOPS (Trace Organics

Platform Sampler) – a better way?

• Pepsi cans – low tech

Colloids• Typically a 0.7 mm filter

is used, which allows small particles to pass through to be quantified with the apparent dissolved phase.

• This leads to the “solids concentration effect”. The apparent distribution between dissolved and particle phases changes as the total amount of solids increases.

log KOW

5.0 5.5 6.0 6.5 7.0 7.5 8.0

log

KO

C5

6

7

87/5/98 sample

With correction:log KOC = 1.06*log KOW - 0.08

Without correction:log KOC = 0.71*log KOW + 1.86

c

iw

isid

C

CK

oc

idioc f

KK

Other sampling

• When sampling for sediment, biota, etc, homogenization and collection of a representative sample are paramount.

• Volatilization still a problem – refrigerate or freeze immediately

Extraction• Techniques:

– Soxhlet Extraction– Accelerated Solvent

Extraction (ASE) (high T and pressure minimize amount of solvent needed)

• Solvents:– Dichloromethane

(toxicity?)

– Pet Ether– Hexane

(leave behind lipids or more polar compounds)

Rotovap

Blowdown

Cleanup• Use column chromatography to remove

interfering compounds from your sample• Type of analytical method determines the

rigor of the cleanup

Adsorbent: Sample Matrix:

Alumina Cleaner Air

Silica Sediment

Florisil Dirtier Sludge

Size Exclusion Chromatography Biota(to remove lipids)

Surrogate recovery

Our Alumina Cleanup

• Bake alumina at 550ºC overnight

• Deactivate with 3% wt water

• Precondition column

• F1 = 13 mL Hexane = PCBs

• F2 = 15 mL 2:1 DCM/hexane = PAHs

• OCPs, PBDEs split between F1/F2

Detection

• Detection method is determined by concentration of compound in environmental matrixes.– PCBs = Electron Capture Detection or High-

Resolution GCMS – PAHs = GCMS EI– PBDEs = GCMS NCI– Cl Pesticides = GCMS NCI– PCDD/Fs = High-Resolution GCMS

GC/ECDInvented by Lovelock around the late 1950s and early 1960s.

Uses a radioactive Beta emitter (electrons) to ionize some of the carrier gas and produce a current between electrodes.

When organic molecules that contain electronegative functional groups, such as halogens, phosphorous, and nitro groups pass by the detector, they capture some of the electrons and reduce the current measured between the electrodes.

The ECD is as sensitive as the FID but has a limited dynamic range and finds its greatest application in analysis of halogenated compounds.

Cost ~ $30,000

GC/MSSolute molecules are ionized in the ion source

Resulting fragments are separated on the basis of their mass/charge ratio then detected by an analyzer unit

Ionization by electrons (EI) or gas molecules (CI, Negative or Positive).

NCI similar to ECD (little fragmentation, best for halogenated compounds)

Scan mode give entire mass spectrum = good for identification of unknowns

SIM (selective ion monitoring) mode = much more sensitive

Cost ~ $100,000

QA/QC

• Sample contamination

• Reproducibility

• Tracking of mass

• Representativeness of samples?

Avoiding Contamination

• Cleanliness– Bake glassware at 450°C overnight– New aluminum foil– High grade solvents– New building!

• Cleaning sampling equipment sometimes difficult– Blanks, blanks, blanks

Reproducibility

• Side-by-side samples

• Duplicates

• Matrix spikes

• Surrogates

Mass Tracking• Surrogates

– Added to track recovery through the various sample processing steps

– Must have same or similar physical-chemical properties as analytes

– Deuterated or 13C labeled– Non-native congeners (PCBs 14, 23, 65, 166)

• Internal standards– Added to allow quantification of mass even though

volume is not known– Deuterated or 13C labeled– Non-native congeners (PCBs 30, 204; BDE 75)

Representativeness of samples?

• Homogenize sediments (Bass-o-matic)

• Take lots of samples

• 12th day sampling

Special considerations for PBDEs

• Flame retardants – designed to break down at high temperatures!

• BDE 209 has 10 bromines– extremely labile– MW = 960 g/mol!

• Use cold on-column injection

• Very short GC column• Avoid light

The Pitfalls of Measuring PCBs by ECD

This is a GOOD chromatogram!

Some PCBs co-elute, and there ain’t nothin’ you can do about it.

Example:

PCB s 110+77

EPA Method 1668A

Uses High-resolution GC/MS (about $1 million)

13C labeled compounds:3 field stds28 surrogates3 cleanup stds5 recovery stds= 39 stds

Trade-Offs

Would you rather have 28 surrogates with 60% recovery, or three surrogates with 95% recovery?

209 congener method can reveal surprises

Contract labs are far from infallible

Contract labs only love your money, not your samples

PCB 11

3,3’-dichlorobiphenyl is not found in Aroclors

NYSDEC “found” it accidentally in effluent from PVSC

It is produced inadvertently during pigment manufacture (see Litten et al., 2002)

Most data sets do not look for it.

Co-elution a problem even for method 1668A

1668A can differentiate between (for example) PCBs 110 and 77, even though they co-elute, because 110 has 5 chlorines and 77 has 4.

BUT PCBs within a homolog group that co-elute are still quantified together

Contract labs report all co-eluting congeners under the congener with the lowest IUPAC number. 

Example: what is reported as PCB 93 is really 93+95+98+100+102 (and is primarily PCB 95).

Cost

METHOD 1668 CALIBRATION SOLUTIONS from Cambridge Isotope Laboratories.

METHOD 1668 CALIBRATION SOLUTIONS SET5X0.2ML $2,400.00 METHOD 1668 DAILY CALIBRATION CHECK STANDARD 0.2 ML $495.00 METHOD 1668A CALIBRATION SOLUTION CS0.2 0.2 ML $625.00 METHOD 1668A CALIBRATION SOLUTION CS1 0.2 ML $625.00 METHOD 1668A CALIBRATION SOLUTION CS2 0.2 ML $625.00 METHOD 1668A CALIBRATION SOLUTION CS3 0.2 ML $575.00 METHOD 1668A CALIBRATION SOLUTION CS4 0.2 ML $625.00 METHOD 1668A CALIBRATION SOLUTION CS5 0.2 ML $625.00 METHOD 1668A CALIBRATION SOLUTIONS CS1-CS5 SET5X0.2ML $2,450.00 METHOD 1668A CLEAN-UP STANDARD SOLUTION 1.2 ML $575.00 METHOD 1668A INJECTION INTERNAL STANDARD SOLUTION 1.2 ML $1,650.00 METHOD 1668A NATIVE TOXICS/LOC SOLUTION 1.2 ML $795.00 METHOD 1668A TOXICS/LOC/WINDOW DEFINING SOLUTION 1.2 ML $2,450.00

Price per sample ~$1,000

The pitfalls of measuring OCPs by ECD

Old Method (ECD): New Method (GC/MS NCI): Similar to EPA Test Method 8080 Based on a method developed by

Shannon Nicole Brown (M.S., U. of Iowa, 2000)

Uses electron capture detection (ECD)

Uses mass spectrometry (MS) with negative chemical ionization (NCI)

6 Pesticides quantified 24 Pesticides quantified

Identifies pesticides based on retention time

Identifies pesticides based on retention time and major ion/secondary ion ratio

If analytes co-elute they will not be quantifiable

Co-eluting "interferences" can be excluded

Jersey City: Gas- phase

y = 0.7574x + 0.1956

R2 = 0.1725

0

5

10

15

20

25

30

35

40

45

0 5 10 15 20 25 30 35 40 45

ECD method

NC

I met

ho

d

N = 28P = 0.03

2,4’-DDT

y = 1.1037x + 3.8032

R2 = 0.79770

10

20

30

40

50

60

0 10 20 30 40 50 60

ECD method

NC

I met

ho

d

4,4’-DDE

N = 28P < 0.0001

Comparison of ECD and NCI data

Because OCPs are abundant enough to be detected by a regular GC/MS instrument, the mass spec method is cost-effective.

Conclusions

• Dirty matrices

• Complex mixtures

• Cost/benefit analyses

• Cleanliness

• Blanks, blanks, blanks

• Trust but verify