One of the most commonly encountered organic testing requirements in the environmental field is the testing of soil and water samples for hydrocarbon contamination.
Given our society’s use of these compounds over the last century and the fact that the world is dependent on them for modern day life, it is not difficult to see the potential for widespread environmental contamination. Although other chemicals may pose greater risks, the sheer volume of oil based products used today dwarfs these. Whether these are volatiles such as benzene in petrol, high banded carbon fractions in diesel or polyaromatic hydrocarbons from sources of combustion, chromatography is a key tool in the identification and quantification of these contaminants.
Petroleum hydrocarbons are a complex mixture of potentially hundreds, if not thousands, of different chemicals covering aliphatic, aromatic, saturated and unsaturated groups. From a human health and remediation perspective, however, it is easier to group these compounds based on their general physiochemical properties and to this end equivalent carbon numbers are used. There are a number of different banding systems employed depending on local requirements, but of these the criteria working group bands (CWG banding) are the most routinely requested.
The start – crude oil and distillation
Crude oil is the parent source of all petroleum hydrocarbons and is the fossilised remains of prehistoric algae and zooplankton which have undergone catagenesis to produce hydrocarbons. It is essentially a complex mixture of hydrocarbons and, as the name suggests, its composition is predominantly carbon and hydrogen. Alongside these two elements one can also expect to find nitrogen, oxygen and sulphur, as well as trace metals.
As a material crude oil is not commercially useable without undergoing additional refining to form the products associated with petroleum hydrocarbons. The refining process, via fractional distillation and cracking (the breaking down of large hydrocarbons), is used to form smaller, more commercially valuable, hydrocarbons.
The starting composition of the crude oil is dependent on where it originated, but fundamentally covers a wide range of compounds from the smallest hydrocarbon compound methane (CH4) all the way through to black tarry substances such as heptacontane (C70H142). As well as compounds with differing chain lengths, the unrefined crude oil comprises a number of different chemical functional groups including saturated and unsaturated straight chain hydrocarbons, saturated and unsaturated cyclic hydrocarbons and compounds containing heteroatomic elements such as nitrogen, oxygen and sulphur present in both straight chain and cyclic forms. These different functional groups, which can be present in almost all of the carbon chain lengths, add to the overall complexity of the crude oil.
“to separate the different compositional compounds into their more commonly encountered products the crude oil is heated in a furnace and then fractionated in a distillation tower”
To separate the different compositional compounds into their more commonly encountered products the crude oil is heated in a furnace and then fractionated in a distillation tower or fractionating column. A distillation tower is effectively a tall column with several condensers coming off at different levels; it is hot at the bottom and cooler at the top. Here the crude oil is separated based on the boiling points of its compositional parts. Compounds that are volatile such as those found in petrol (benzene for example) are only condensed towards the top of the distillation tower and are known as light distillates, whereas those with higher boiling points condense more easily around the middle of the distillation tower – these are known as middle distillates and cover products such as diesel and kerosene. Those compounds remaining after fractionating are heavy distillates and cover products such as lubricating oils, waxes and tar. These may be used ‘as is’ or they may be further processed by cracking to break them down into more valuable, lighter weight distillates.
Total Petroleum Hydrocarbons – TPH
Distillation and refining produce separate, purified products which are then useable as commercial feedstocks for petrol, diesel or any other petrochemical application. Although purified, the products still contain a very large number of individual compounds, each with their own different chemical properties.
As the refined products contain such a wide range of chemicals types and classes it is imperative to be able to resolve these to enable a proper assessment of risk to be undertaken. The primary split for hydrocarbons is to separate them based on their structure, namely aliphatic vs. aromatic, followed by subdividing these into carbon bands. This approach has been widely accepted for use in human health risk assessment (e.g., TPH CWG, Volumes 1 to 5 and Environment Agency (2005a) and is useful when considering the risks to controlled waters, since the fractions can be assigned representative fate and transport properties.
Petroleum products such as petrol or diesel, often referred to as petrol range organics or diesel range organics (PRO and DRO), are examples of different TPH fractions but they do not define what TPH is. As a guide they outline a group of likely contaminants but they don’t identify any of the broader specifics useful for environmental or human health risk assessment. The situation becomes more complex once environmental effects have taken their course – effects such as weathering, leaching, and biodegradation all impact the source material and can significantly alter the material found when compared to the original source material.
In the simplest definition, aliphatic hydrocarbons are those that contain carbon and hydrogen joined together in straight chains, branched chains or non-aromatic rings. Aliphatic compounds may be saturated (e.g. octane) or unsaturated (containing double bonds e.g. octene). The simplest aliphatic hydrocarbon is methane, CH4. Aromatic hydrocarbons on the other hand are always ring structures and are unsaturated. The simplest aromatic hydrocarbon is benzene, C6H6. Aromatic hydrocarbons may comprise multiple rings fused together such as the polyaromatic hydrocarbons (PAHs) and/or may contain aliphatic chains connected to the aromatic ring, e.g. ethylbenzene or methyl naphthalene.
The number of possible aliphatic/aromatic compounds is effectively limitless. The analytical laboratory can analyse for a significant number of these as individual compounds, but it is neither sensible nor practicable to analyse for every compound this way. In addition the interpretation of such data from a remediation or human health perspective would be arduous at best hence the adoption of aromatic/ aliphatic splits (aliphatic/aromatic splits) and carbon banding.
“the number of possible aliphatic/aromatic compounds is effectively limitless”
Whilst techniques such as Infra-Red (IR) or gravimetric analysis can provide ‘total’ concentrations in relation to environmental contamination they do not provide enough detail to evaluate the potential risks. Two sites may both contain 5000 mg/kg of TPH, but if one is predominantly large paraffinic hydrocarbons and the other coal tar loaded with Benzo(a)pyrene, the risks – and the remediation approach – are going to be significantly different.
Aromatic and aliphatic splits
Depending on the analytical operation of the laboratory there are two main options for separating aromatic and aliphatic hydrocarbons. The most established is the separation post extraction using an off-line solid phase extraction (SPE) approach.
For aliphatic/aromatic splits using SPE an aliquot of extract is passed through a preconditioned SPE cartridge which separates the polar aromatic compounds from the non-polar aliphatic ones. The interaction of the aromatic compounds with the SPE material means that these compounds are retained on the surface of the SPE material whilst the non-polar compounds pass through unhindered. By changing the polarity and strength of the solvents used (for example from hexane to dichloromethane), the aromatic compounds can then be eluted from the SPE cartridge and collected separately. The two fractions are then concentrated and analysed.
Whilst cheap and simple to setup and operate, there are drawbacks that the laboratory operative needs to be aware of. These include breakthrough where the aromatic fractions bleed in to the aliphatic fractions and also variations in the ability of the SPE cartridges to separate the two fractions.
A more modern take on this separation is to use a technique known as GCxGC. This technology is an evolution over standard GC, but rather than using one analytical column for separation, two are used interfaced with a modulating device controlling column flow to the detector. Chromatographic conditions are based on normal single column 17 Chromatography programmes with a non-polar column in the first dimension providing a comparable separation to existing TPH methods. The second column provides a polar separation which resolves the aliphatic components from aromatic (polar) compounds.
The main advantage of GCxGC is that no physical split is required; therefore no losses or breakthrough can occur. This results in more consistent and reliable data.
Carbon bands and equivalent carbon numbers
To fully understand carbon banding, one needs to understand how the analysis of hydrocarbons is undertaken in the laboratory using gas chromatography. The specifics of a certain extraction and the pros and cons of one approach versus another are not important at this stage as the fundamentals are the same.
Gas chromatography (GC) is essentially the separation of a complex mixture of compounds into it constituent parts. When connected to a detector such as a flame ionising detector (FID) or mass spectrometer (MS) the analyst has the ability to separate complex mixtures and then quantify the amount of each component. The complex mixture in this example would be the extracted TPH injected into the instrument.
GC uses a glass column and a combination of temperature and an inert gas to separate compounds based on their boiling points (other principles exist for separation but this is the simplest and most usually encountered). In this regard the separation is very similar to the fractional distillation that generated the products from the original crude, except that rather than being condensed for further purification, the products are detected and a chromatogram produced.
“in the GC, the most volatile compounds elute from the column first and appear on the left hand side of the chromatogram”
In the GC, the most volatile compounds elute from the column first and appear on the left hand side of the chromatogram. As the temperature increases over time, the heavier compounds elute and are detected. By using a calibration solution containing known carbon compounds (e.g. decane (C10), hexadecane (C18), and pentatriacontane (C35)) specific carbon ranges can be visually seen on the chromatogram and reported. In practice a mixed n-alkane standard covering all even numbered, straight chain, alkanes from C8 to C36 is often used allowing different regional or customer specific carbon ranges to be reported. The separation of alkanes by boiling point via GC also opens up the possibility for classifying non-alkane hydrocarbons using a technique known as equivalent carbon numbers or EC.
The EC defines where certain bands start and finish relative to each specific n-alkane. This is a simple and practical solution to an otherwise complex problem where classification of complex mixtures is required and as such has found use worldwide. A number of different carbon band groupings are used but the most frequently requested at the laboratory is the CWG or criteria working group bands shown in the table below.
As can be seen the CWG and Environment Agency bands are very similar but with the UK agency adding supplementary bands covering >C35 to C44, split into aliphatics and aromatics, as well as a total for >C44 to C70.
The table shows the bands using a ‘greater than’ sign (>) for example >C8-C10. Although often seen in the format C8-C10 the correct term would be >EC8-≤EC10 i.e. greater than C8 but less than or equal to C10.
By definition the EC for aliphatics directly relates to the carbon number, e.g. octane (C8H18) has an EC of eight, but for aromatics this can only be determined empirically. Naphthalene and Benzo(a)pyrene are polyaromatic hydrocarbons with the structures C10H8 and C20H12 respectively. Although chemically they have 10 carbons and 20 carbons apiece, their equivalent carbon numbers are EC12 and EC31.1
In summary, the use of carbon banding with aliphatic/aromatic fractionation provides information on the distribution of hydrocarbons within a mixture in terms of their size and their physiochemical properties. This can be used to characterise a hydrocarbon mixture of unknown origin and aids in the identification of the likely product, or mixture of products. The analysis can also be used in an assessment of the risks posed to controlled waters.
Together with the analytical constraints of the laboratory methods there are a number of classifications that need to be employed in order for a more consistent, universal, approach to petroleum hydrocarbon analysis. The ‘Total’ in TPH as a moniker is too broad and as a test is essentially meaningless, unless defined further by a carbon range and ideally also by a further classification as to the exact type present e.g. aliphatics vs. aromatics as well as the analytical technique used.
Owing to the complex composition of TPH a change in analytical procedure from one lab to another can result in a step change (up or down) in reported results. The difference between GC-FID and GC-MS for example or manual separation of aliphatic/aromatic splits vs. GC x GC will inevitably change the result.
“different proponents of on-site monitoring will proclaim of the benefits of instant results and low costs, but whilst useful for preliminary screening it cannot provide the accuracy of an accredited laboratory”
In addition to variabilities within the laboratories there is a growing trend for onsite monitoring. This is a growing area and whilst not in its infancy it is not yet widely used in the UK. Different proponents of on-site monitoring will proclaim of the benefits of instant results and low costs, but whilst useful for preliminary screening it cannot provide the accuracy nor level of detail an accredited laboratory with an experienced operative can.