Chemical Fingerprinting

Environmental professionals often need to do more than locate and clean up pollutants. They may need to help identify the sources of the contamination. Someone always has to pay for a remediation effort. The high costs that are often involved mean that any prudent potentially responsible party (PRP) will want convincing proof that he or she must accept responsibility for the pollution. Polluted sites often have a history of having different owners and uses, any or all of which may have contributed to the present site contamination. Off-site sources may also be responsible for all or part of the current pollution problems. Allocating responsibility for the expenses involved in remediation often becomes a legal issue. The methods used in building a legally defensible case that assigns responsibility for soil and groundwater pollutants have become part of a branch of environmental investigations known as environmental forensics.

One of the most common applications of environmental forensics is identifying the sources of fuel contamination. The use of petroleum fuels has been increasing for the past 100 years and fuel contaminants are pervasive wherever they have been used. There are many potential sources of fuel pollutants, from numerous small gasoline stations to large refinery operations, roadway accidents to criminal acts of disposal, pipeline breaks to tank corrosion. It is no wonder that the origin of any particular site contamination may often be legitimately questioned. Chemical fingerprinting of fuels is one tool in the toolbox of chemical forensics.

In chemical fingerprinting, the chemical characteristics of fuel contaminants are used to help distinguish among different possible sources of the pollutants. For example, knowing that only gasoline was ever used on a certain site, but that the groundwater hydrocarbon contaminants found there are from diesel fuel, clearly points to an off-site source.

First Steps in Chemical Fingerprinting of Fuel Hydrocarbons

Successful chemical fingerprinting of hydrocarbon contamination frequently consists of the sequential application of several investigative steps:

• Identify the fuel type of the contaminants. Are they gasoline, diesel fuel, or fuel oils? Each of these classes has unique chemical characteristics.

• If possible, distinguish between the ages of different contaminated samples. Weathering processes often introduce predictable changes in the chemical makeup of fuels. If samples have weathered differently because of different exposure times, they will have different chemical profiles. Certain compounds in fuel samples are more readily biodegraded, solubilized, or volatilized than others. This causes changes in the overall fuel composition as time passes. The exact nature of the changes is always site specific, but sometimes the ages of different samples can be clearly demonstrated to be significantly different. In some cases, the presence of certain degradation products may be a useful indicator of age. Also, fuel compositions have changed historically, as more efficient refinery production practices were developed and as clean air legislation and automotive engine development mandated changes. For example, the elimination of lead-based octane boosters in gasolines began in the early 1980s. The presence of lead may indicate a vintage rather than recent spill, especially if accompanied by dichloroethane and/or dibromoethane, which were lead scavenger gasoline additives.

• Look for unique chemical compounds that can serve as "markers" which may be present in contamination from one source, but not from another. An example might be the presence of MTBE or ethylene dichloride in some fuel-contaminated samples but not in others. Different production practices at different refineries sometimes produce subtle but distinctive differences in their fuel products, which can help to distinguish between possible sources. This approach generally requires extensive knowledge about the chemical composition of contamination at the site and, sometimes, knowledge about the chemicals previously and currently used at the site.

Chemical fingerprinting is sometimes fairly simple and sometimes very complicated, even impossible. A few approaches are described here to indicate the possibilities. Complicated cases will require the help of experienced forensic chemists.

Identifying Fuel Types

The most useful analytical tool for identifying different fuel types is the gas chromatograph (GC), especially with a mass spectrometer detector. Such instruments are referred to as GC/MS. When fuels are analyzed in a gas chromatograph, retention times of different hydrocarbon compounds closely correspond to their boiling points; the higher the boiling point, the longer the retention time. Thus, lower boiling point gasoline components elute from the gas chromatograph column earlier than higher boiling point diesel components.

Physical and chemical characteristics of different fuels were described in Section 5.2. Table 5.1 and Figure 5.2 show that carbon number ranges characteristic of different fuel types correspond to different boiling point ranges. In general, the more carbons in a petroleum molecule, the higher is its boiling point (see the discussion of London forces in Chapter 2). In a gas chromatogram of fuel hydrocarbons, the longer the retention time of a peak, the higher the boiling point of the corresponding compound and the more carbons in the molecule. Although the presence of structural differences, such as the presence of carbon side chains, introduces some variability, this general principle remains useful. Gas chromatograms of different types of fuel are different from one another and are useful for identifying fuel types.

In Figure 5.13, note that the gasoline signature contains more lightweight components (peaks farther to the left, indicating fewer carbon atoms) than do diesel fuel or lubricating oil. When fuel hydrocarbons are weathered by volatilization, dissolution, and biodegradation, the lighter components on the left side of the fresh gasoline and diesel fuel signatures are lost first from the free product. Thus, weathered diesel does not resemble gasoline because it lacks the lightweight components that are characteristic of gasoline. Likewise, weathered gasoline does not resemble diesel fuel because it lacks the heavier components that are characteristic of diesel fuel.

FIGURE 5.13 Gas chromatograms showing the differences in chromatographic signatures between different types of fresh and weathered petroleum hydrocarbon free product. Figure (f) at lower right is a GC of free product containing a mixture of gasoline, diesel fuel, and motor oil. Humps, where the chromatogram rises above the baseline, are due to hundreds of different hydrocarbon compounds, which are not chromatographically resolved.

FIGURE 5.13 Gas chromatograms showing the differences in chromatographic signatures between different types of fresh and weathered petroleum hydrocarbon free product. Figure (f) at lower right is a GC of free product containing a mixture of gasoline, diesel fuel, and motor oil. Humps, where the chromatogram rises above the baseline, are due to hundreds of different hydrocarbon compounds, which are not chromatographically resolved.

A simple and common indicator of the age of gasoline spills is the ratio of benzene to other components of the fuel. Benzene is more soluble, volatile, and degradable than other common gasoline constituents and is depleted relative to other compounds in samples with older contamination.

Rules of Thumb for Estimating the Age of Gasoline Contamination

Current gasolines contain about 2% to 5% of benzene. Older gasolines may have as little as 1%. EPA promotes the reduction of benzene to 1% or less for health reasons. Because benzene is the most soluble of the BTEX group

1. A sample from water in contact with fresh gasoline will have a benzene-to-toluene ratio of about 0.6 to 1.5.

2. A gasoline-contaminated water sample containing much less benzene than toluene (benzene/toluene < 0.1) indicates older gasoline contamination, perhaps 6 months to 2 years old.

3. A water sample where the concentration ratio of benzene-to-total BTEX is less than 0.001 may be 5 to 10 years old, or even older.

4. Another clue to the age of gasoline contamination is based on tetraethyl lead, organic manganese compounds, ethylene dibromide (EDB), and ethylene dichloride (EDC), which were common additives to gasoline before 1980. Therefore, the presence of these compounds is supporting evidence for gasoline contamination originating before 1980. EDB and EDC, however, are present in some agricultural chemicals and must be used as an indicator with caution.

TABLE 5.5

Approximate Composition of Fresh Diesel Fuel (C-6 to C-24)

Class of Compound Percent

N-alkanes (degrade fastest; smaller alkanes degrade faster than larger alkanes). 40%

Iso- and cyclo-alkanes (degrade slower). 36%

Isoprenoids (degrade very slowly). 3-4%

Aromatics (most soluble; mainly parent and alkylated benzenes, naphthalenes, hydrindenes, phenanthrenes, 20% and fluorenes).

Polar (water soluble sulfur, nitrogen, and oxygen compounds). 1%

Age-Dating Diesel Oils

Normal alkanes (linear hydrocarbons) in the approximate carbon number range C-6 to C-24 are the most abundant components of fresh diesel oils — although many other types of hydrocarbons are also present (see Table 5.5). The table also shows that the different types of hydrocarbons in diesel oils biodegrade at different rates. Isoprenoids are branched, unsaturated hydrocarbons that biodegrade more slowly than linear alkanes with similar masses because their chemical structure inhibits biodegradation.

The solubilities and volatilities of oil range n-alkanes and isoprenoids are quite low and similar for the two chemical classes. Therefore, they have similar rates of weathering by nonbiological processes. However, because isoprenoids biodegrade much more slowly than w-alkanes, the abundance ratio of w-alkanes to isoprenoids changes over time where biodegradation is occurring. Because the passive biodegradation rates of diesel in soils are fairly uniform at similar sites,5 the corresponding changes in composition can be used for estimating the age of the diesel contamination.

In 1993, Christenson and Larsen5 found, as others had, that by comparing gas chromatograph peak-height ratios of pristane (C-19 isoprenoid) with w-heptadecane (linear C-17 alkane) an estimate can be made of the degree of biodegradation that the sample has undergone. The ratios of phytane (C-20 isoprenoid) and w-octadecane (linear C-18 alkane) can be compared in a similar way. Using the relative extents of biodegradation, the relative ages of fuel contamination in different samples from similar sites can be estimated.

When the composition of fresh diesel fuel is compared with weathered fuels that have biodegraded in the subsurface environment, certain composition changes are apparent. In particular, w-alkanes dominate the composition of fresh fuels and isoprenoids dominate the composition of highly degraded fuels. The C-17/pristane peak-height ratio falls from about 2 (for fresh diesel) to 0 (after about 20 years). Figure 5.14 compares gas chromatograms of fresh and biodegraded No. 2 diesel oil. Changes in the relative peak heights of w-alkanes and isoprenoids are readily apparent.

Christenson and Larsen5 also found a linear relation between the age of diesel oils and the C-17/pristane peak-height ratio. From their data, it is possible to determine the age of a diesel spill to within about 2 years for diesel that is between 5 and 20 years old, that was created by a single sudden spill event, and that has not been "weathered" significantly except by biodegradation. A linear best-fit to their data yields the equation

Age of diesel fuel (yr) = -8.3(C-17/pristane ratio) + 19.5. (5.5)

Example 5.4: Fingerprinting Fuel Contaminants at an Industrial Site

The following GC/MS data, taken from soils at an industrial site with contamination by gasoline, diesel, and heavy oil hydrocarbons, indicate that at least two different diesel spill events occurred, separated by approximately 6 to 10 years. The mass spectrometer detector was sometimes operated in the single-ion mode, which increases sensitivity. Single-ion monitoring consists of leaving the mass spectrometer tuned to a fixed mass number as the gas chromatogram peaks elute, rather than continually scanning the entire mass range for each GC peak. The difference between single-ion and full-range monitoring is shown in Figures 5.15a and 5.15b.

In order to estimate the age of the diesel fuel shown in Figure 5.15, the GC/MS spectra in the pristane/phytane region (15-17 min retention time) must be expanded. Expanded single-ion spectra are shown in Figure 5.16 for samples from Boreholes A and B, which are separated by about 200 yd.

Pristane, at 16.06 min in Figure 5.16a, biodegrades much more slowly than the C-17 ^-alkane at 16.00 min. The same is true for phytane, at 16.92 min, and the C-18 n-alkane at 16.83 min. The retention times for these peaks are slightly different in Figure 5.16b.

In the Borehole A sample, the C-17/pristane peak-height ratio is 0.33 and the C-18/phytane ratio is 0.26, indicating significant biodegradation. Although, site-specific conditions will determine how much time this amount of degradation would require, the diesel in this sample is at least 15-20 years old. Equation 5.4 indicates an age of about 19 years.

In the Borehole B sample, the C-17/pristane ratio is 1.2, which indicates a more recent diesel spill. Equation 5.4 suggests that the diesel fuel from Borehole B is about 9 years old.

Rules of Thumb

1. Fresh diesel contains more n-alkanes than isoprenoids, such as pristane or phytane. Therefore, if relative depletion of the n-alkanes is observed, it indicates that biodegradation has occurred.

2. In a fresh diesel fuel, the C-17/pristane and C-18/phytane ratios are close to 2. Any ratio less than about 1.5 indicates that biodegradation has occurred.

FIGURE 5.14

(a) A gas chromatogram of typical fresh No. 2 diesel oil. The numbered peaks are linear ^-alkanes, where the number represents the carbon number (e.g., C-17) of the alkane. The ^-alkanes are the most abundant compounds in fresh diesel fuels and dominate the composition. The peaks of several isoprenoids are labeled. The peak-height ratio of C-17/pristane is about 2:1. (b) A gas chromatogram of biodegraded No. 2 diesel oil. Isoprenoids are more abundant than ^-alkanes. The peak-height ratio of C-17/pristane is about 0.8, indicating an age of about 13 years according to Equation 5.5.

FIGURE 5.14

(a) A gas chromatogram of typical fresh No. 2 diesel oil. The numbered peaks are linear ^-alkanes, where the number represents the carbon number (e.g., C-17) of the alkane. The ^-alkanes are the most abundant compounds in fresh diesel fuels and dominate the composition. The peaks of several isoprenoids are labeled. The peak-height ratio of C-17/pristane is about 2:1. (b) A gas chromatogram of biodegraded No. 2 diesel oil. Isoprenoids are more abundant than ^-alkanes. The peak-height ratio of C-17/pristane is about 0.8, indicating an age of about 13 years according to Equation 5.5.

FIGURE 5.15 Full-range and single-ion GC/MS spectra of a contaminated soil sample from Borehole A. (a) GC/MS full-range spectrum of fuel contamination extracted from soil at Borehole A. The distribution of compounds indicates that the soil contains gasoline, diesel, and heavy oil hydrocarbons. (b) Single-ion spectrum at mass 71 of the same sample. Mass 71 is a fragment ion common to many hydrocarbon compounds of C-7 and larger. It is useful for diesel fingerprinting because the mass 71 ion is produced abundantly in the fragmentation of pristane and phytane.

FIGURE 5.15 Full-range and single-ion GC/MS spectra of a contaminated soil sample from Borehole A. (a) GC/MS full-range spectrum of fuel contamination extracted from soil at Borehole A. The distribution of compounds indicates that the soil contains gasoline, diesel, and heavy oil hydrocarbons. (b) Single-ion spectrum at mass 71 of the same sample. Mass 71 is a fragment ion common to many hydrocarbon compounds of C-7 and larger. It is useful for diesel fingerprinting because the mass 71 ion is produced abundantly in the fragmentation of pristane and phytane.

FIGURE 5.16 Expansion of mass 71 single-ion spectra in the pristane/phytane region of contaminated soil from Boreholes A and B. (a) Expansion of the mass 71 spectrum in the 15-17 min retention-time region for the Borehole A sample. The C-17/pristane peak-height ratio is about 0.26 indicating significant biodegradation. (b) At another location on the site, Borehole B, the C-17/pristane peak-height ratio is about 1.2 indicating diesel contamination that is younger and much less biodegraded.

FIGURE 5.16 Expansion of mass 71 single-ion spectra in the pristane/phytane region of contaminated soil from Boreholes A and B. (a) Expansion of the mass 71 spectrum in the 15-17 min retention-time region for the Borehole A sample. The C-17/pristane peak-height ratio is about 0.26 indicating significant biodegradation. (b) At another location on the site, Borehole B, the C-17/pristane peak-height ratio is about 1.2 indicating diesel contamination that is younger and much less biodegraded.

Simulated Distillation Curves and Carbon Number Distribution Curves

A simple approach to chemical fingerprinting is to generate simulated distillation curves (SDCs) and carbon number distribution curves (CNDCs). These plots allow a relative comparison of the concentrations of volatile and semivolatile compounds that are present in a sample, without requiring a detailed analysis of each compound. The curves can be generated automatically from computerized gas chromatogram data without identifying the individual peaks.

A simulated distillation curve shows the percentage of the sample that would be volatilized at various temperatures. Since the volatility of a petroleum compound is closely related to the number of carbon atoms in that compound (compounds with fewer carbon atoms are lighter in weight and boil at lower temperatures than compounds with more carbon atoms), one can estimate the general chemical makeup of petroleum contaminants from an SDC. A carbon number distribution curve gives similar information but is often easier to interpret.

SDC and CNDC curves allow a visual comparison of the mass distribution of chemical compounds that are present in the analytical samples, based on their boiling points and masses. The shapes of these curves are distinctly different for different types of hydrocarbon mixtures. Gasoline, for example, contains relatively high concentrations of lightweight hydrocarbons such as benzene, while diesel fuel normally has very low concentrations of these lightweight compounds, but higher concentrations of heavier compounds. Weathering of organic compounds produces predictable changes in the shapes of SDCs and CNDCs.

When CNDCs and SDCs have distinctive shapes, they can be used as a "fingerprint" for determining whether contamination at one location is different from or similar to contamination at another location. Sometimes this preliminary analysis is sufficient for identification purposes, as in Example 5.5. In other cases, it serves to guide further study.

Example 5.5: Fingerprinting Diesel Contamination With Simulated Distillation and Carbon Number Curves

Mr. A, the owner of a newly purchased mountain home, frequently, but not always, detected strong fuel odors in his basement shortly after diesel fuel was delivered to a neighbor's underground storage tank (UST). Mr. B, the neighbor, was cooperative and allowed Mr. A to take samples from the UST for analysis and comparison with soil samples from around Mr. A's new home.

In Figure 5.17a, the simulated distillation curve for No. 2 diesel fuel from Mr. B's UST lies somewhat to the left of the laboratory standard for diesel, showing that it contains a higher percentage of lower boiling (lighter weight) compounds. This could indicate that the UST fuel was contaminated slightly with gasoline (perhaps from a tanker truck used to carry both types of fuel), or that it was specially formulated for use in cold weather. In Figure 5.17b, the carbon number distribution curve for the UST fuel has a distinctive fingerprint that includes two prominent peaks at C-11 and C-13.

Figures 5.18a and 5.18b show simulated distillation and carbon number distribution curves for a soil sample collected from a free product seep adjacent to the foundation of Mr. A's home. Not only is the SDC for the foundation seep sample very similar to the SDC for the UST sample, but also its CNDC contains the two distinctive peaks at C-11 and C-13 that were also seen in the UST sample.

Mr. B acknowledged that this data strongly implicated his UST to be the source of contamination at Mr. A's home, even if it did not explain why the leak appeared to be erratic. When his tank was leak tested, it was found that the filler pipe had cracked where it was fastened to the tank below the soil's surface. It leaked only when the tank was overfilled and diesel rose in the fill pipe above the crack.

Healthy Chemistry For Optimal Health

Healthy Chemistry For Optimal Health

Thousands Have Used Chemicals To Improve Their Medical Condition. This Book Is one Of The Most Valuable Resources In The World When It Comes To Chemicals. Not All Chemicals Are Harmful For Your Body – Find Out Those That Helps To Maintain Your Health.

Get My Free Ebook


Post a comment