Trace Organic Analysis

Frequently Asked Questions

	What is an organic compound?		What are separations?
	What is trace organic analysis?		What are detection techniques?
	What are units of measurement?		How do you do trace organic analysis?
	What is the central role of mass spectrometry in environmental analysis?		What are solvent properties for chromatography and sample handling?
	How do you approach an analysis problem?		How do you interpret mass spectra?
	How do you acquire and process mass spectrometry data?		What are the basic equations of capillary electrophoresis?
	What are the basic equations of chromatography?		How do you calculate isotope abundances?
	What are extraction techniques?		What is solid phase extraction?
	What is derivatization and what role does it play in analysis?		What are EDCs?

For additional information regarding organic chemistry nomenclature and numerous aspects of analytical chemistry, see the web links at: http://www.epa.gov/nerlesd/chemistry/ppcp/relevant.htm#Chemistrylinks.

What is an organic compound?
An organic compound (made up of one or more atoms of different types) always contains carbon and other elements such as hydrogen, oxygen, nitrogen, and sulfur. If the compound contains a metal or a nonmetal such as Se, then the compound is generally regarded as an organometallic compound. Methane (CH₄) is considered an organic compound but carbon dioxide (CO₂) is not. Thus, organic compounds usually contain hydrogen as well as carbon. Organic chemistry is the study of the chemistry of carbon (or organic compounds as we have defined them).

What is trace organic analysis?
Generally, we will define trace organic analysis as analyses for organic compounds present at or below the part-per-thousand level in a sample. The analysis provides the identity (qualitative result) and the amount (quantitative result). The challenge is to perform the analysis under conditions where the relative amounts of other substances in the sample are vastly greater than that of the analyte. Often, we will find it necessary to separate the analyte from the bulk of the other substances (potential interferences) in the sample (matrix) before a final determination (amount) can be accomplished. Generally, the final determination involves a final separation/detection. For an operational definition of trace organic analysis (real application) see How do you do trace organic analysis?

What are units of measurement
Environmental measurements are often reported in dimensionless units such as part per million (ppm). If 1 liter (1.06 quarts) of water were involved, then 1 ppm would amount to 1 milligram (mg) per kilogram (kg) (1 L). This could also be expressed as 1 microgram per g. Other commonly used units include parts per billion (1 microgram per kg) and parts per trillion (1 nanogram per kg).

The gram is a basic unit of mass mesurement that corresponds to the mass (weight) of one milliliter (mL) of water (a mL is about 1/1000th of a quart). The units of scale then go as powers of ten such that 1000 mL equals 1 liter (L), but 1/1000th of a mL is a microliter (1 µL). The same powers relation is used for the gram scale; thus, 1000 g is 1 kilogram (kg) and 1/1000th of a gram is a milligram (mg) and 1/1000th of a mg is a microgram (µg).

Concentration units (molar, M) may also be employed based on moles per liter (M) of a compound. The moles unit is dimensionless and is calculated by dividing the grams present by the gram molecular weight (the molecular weight expressed in grams) of the compound. The reason for the usefulness of this number is its ability to refer to a known number of molecules per unit volume rather than a weight per unit volume. Thus if a compound is present in 1 L of water at 1 M, then there are 6.023 X 10²³ molecules of the compound present in the liter. In 100 mL, there would be 6.023 X 10²² molecules present, a still formidable number. (See: Environmental Chemistry:  Measurement - Methods - Quality Assurance - Statistics.)

What are separations?
Separations are physical methods used to divide a sample into discrete components over a time or distance scale from an arbitrary starting time or distance. Thus, for example, in a horse race all horses start even and together. At the end of the race after circling the track they cross the finish line at different elapsed times and are also physically separated from one another. Many separation techniques work similarly by having the mixture start at the beginning of a column of separating material and components of the mixture exit the column one at a time as some flowing material (liquid or gas) helps to sweep them along. The ideal separation is fast, efficient (narrow peaks), and separates everything (selective). In complex real samples the ideal separation is rarely, if ever, achieved. Gas chromatography is a favorite separation technique in environmental analysis. Other techniques include high performance liquid chromatography, capillary electrophoresis, and thin-layer chromatography. A variety of animations are available on the web and which portray the principles of separation and detection. See "Chemistry-Based Animations" at Chemistry Links.

	Gas Chromatography (GC) In gas chromatography a volatilized compound is carried through an open tubular column by the flow of an inert gas. During its transit through the column, it interacts with a thin film that coats the inside of the column, partitioning into that phase and then partitioning back into the gas phase continuously. Differences in affinity for the stationary phase (coating on the walls) among substances accounts for the separation resulting in different retention times for each substance. An example of a separation is given in Figure 1 for some acidic herbicides that have been methylated to make them more chromatographable by GC/MS. The peaks are narrow and the selectivity is good. Figure 1
	High Performance Liquid Chromatography (HPLC) HPLC has some similarities to gas chromatography but is carried out in the liquid state and does not require volatilization of compounds. In this instance, the mobile phase (flow by pressure) is a liquid solvent or solvent mixture and the substances partition between this mobile phase and a stationary phase that varies among several different materials. Traditional HPLC used a "normal" phase mode where the mobile phase was hexane and mixtures of hexane with a more polar solvent such as ether. The stationary phase was silica. In the most popular format today called "reverse" phase, the mobile phase is made up of water and an organic solvent such as acetonitrile or methanol. The stationary phase is often a derivatized silica containing octadecyl(C18)silyl bonds to the silica surface (silanol groups). Figure 2 illustrates an HPLC separation with diode array detection (DAD) of acidic herbicides on C18. Figure 2
	Capillary Electrophoresis (CE) Capillary electrophoresis is a technique that effects separations of ions by their mobility in an electrical field (also called free zone electrophoresis). The capillary is filled with an appropriate buffer, a sample is introduced, and ions migrate according to their own mobility superimposed onto the flow of the bulk buffer through the capillary (called the electroosmotic flow, not a pressure-based flow). There are a number of variations to the experiment some of which allow the separation of neutral molecules by the use of micelles. Figure 3 illustrates the separation of salicylic acid and an internal standard prior to determination in effluent using laser-induced fluorescence (LIF) detection at 244 nm excitation. Figure 3
	Mass Spectrometry/Mass Spectrometry (MS/MS) The topic is mentioned here because separations can be performed in dimensions other than temporal (time based as above). In this case we can use gas phase ion chemistry to separate ions in the first mass spectrometer according to their mass to charge ratio (m/z), and then do collision activation of the selected ion to form the separate product ions from the activated ion. Thus, a selectivity is introduced via the m/z ratio and appears on a mass to charge scale. Figure 4 shows a schematic on MS/MS. Figure 4
	High Resolution Mass Spectrometry This topic introduces a second mass to charge approach that distinguishes ions that are only slightly different in mass (assuming an electronic charge of 1). Selectivity is introduced by the narrow ranges of mass allowed through the entrance and exit slits of a double focussing instrument, for example. Figure 5 shows the mass profile of an ion that is close to m/z 181.007. The profile is generated from small voltage jumps that correspond to about 0.0018 u differences. Since the peak is gaussian and has a sigma of about 0.036 u, ions whose m/z values differ by 0.036 are easily discerned. Resolution corresponds to 10000 (m / D m) (181. / 0.0018) See Mass Spectrometry. Figure 5

What are detection techniques?
Substances must somehow be recognized in the sample or in the separation technique applied to the sample. The process is called detection and can be based on a great variety of chemical and physical means. Light is often used to detect compounds because different compounds absorb light differently, and these differences can be measured. Thus, measurements are involved in this stage of the analysis where results are reported in numerical format with appropriate units.

Mass spectrometry is a favorite technique of detection because of its specific, sensitive, and quantitative nature. Additional detector examples include flame ionization, electron capture, infrared absorption, fluorescence intensity, and electrical conductivity. Mass spectrometry itself depends on making and separating ions from the sample (based on one of several ion optics designs) and detecting them via an electron multiplier or photomultiplier detector.

	Electron Multiplier The electron multiplier (Figure 6) uses an initial emission of electrons either from a photocathode or as a result of a collision with an ion and a special metal plate. The electrons cascade through a series of voltage drops such that a multiplication factor results since there is a yield of more than 1 electron for each electron accelerated per stage. Gains of 106 are typical. Figure 6
	Photomultiplier Tube The photomultiplier (Figure 7) uses an intial emission of electrons from a photocathode to produce electrons which then get amplified in an electron multiplier. Gains of 106 are typical. Figure 7
	Diode Array Detector (PhotoDiodeArray) Figure 8 gives a simplified diode array detector configuration. Light is passed through the sample and then dispersed into its various wavelengths and detected by individual photodiode detectors. Small differences in the absorption of the light can be detected and amplified to give an absorption spectrum (full scan) or response on a given wavelength. Figure 8
	Fluorescence Detector A photomultiplier tube (Figure 9) detects fluorescent light from a sample and records the intensity. Fluorescent light is emitted from samples that can fluoresce when illuminated with light of the appropriate wavelength (fluorescent light is shifted to longer wavelengths than the excitation light). Fluorescence detection has the ability to detect extremely low levels of light and accordingly very sensitive molecules may be detected at zeptomole concentrations (10-21 M). Figure 9
	Electron Capture Detector (ECD) For compounds containing halogen (F, Cl, Br, I) or other special functional groups (e.g., NO2) the ECD (Figure 10) is very sensitive. A standing current resulting from ionization of gas by a Beta emitter (e-) is continuously monitored and changes are detected as they result from compounds entering the detector cell. Figure 10
	Mass Spectrometry (MS) The workhorse detector for environmental applications as a result of its sensitivity, specificity, and quantitative nature. Compounds introduced into the instrument at the ion source (instrument internally under vacuum) need to be ionized (or preionized for some liquid introductions) and are then subjected to various ion optics depending on the type of analyzer (time-of-flight, quadrupole, sector, and ion-trap, to name a few) used to separate the ions on a mass-to-charge ratio basis. In Figure 11 a triple sector instrument is illustrated with EBE geometry, representing an electric sector, a magnetic sector, and another electric sector. The ionization process may induce fragmentation or fragmentation may be induced by collisonal activation (collisions between ions and a gas introduced along the ion beam at some point) from which structural specificity may be inferred. Ions are usually detected by electron multipliers or photomultipliers. The basic information in a mass spectrum consists of the m/z ratios corresponding to the various fragments or whole molecular ions and their relative abundances. Data concerning these mass spectra form the backbone of mass spectral libraries where spectra obtained can be searched against what is in the database. Nowadays, spectral libraries containing more than 275,000 spectra are available. Regulatory Mass Spectrometry   In the regulatory setting, mass spectrometry is sought to provide an unequivocal identification of a compound. Issues arise related to what constitutes an identification under the particular context of the analysis. Because full scan EI spectra provide a sufficient identification together with a retention time under most conditions, questions arise usually in the context of ion monitoring or non-EI ionization conditions where few diagnostic ions are present. Full scan EI spectra are mass spectra produced by electron ionization of molecules introduced into the ion source and which, typically, exhibit ions (mass / charge) from the molecular ion region (representing the entire molecule minus 1 electron) down to mass / charge of about 15; fragmentation occurs because the ionization event is relatively energetic and imparts sufficient energy into the resulting ion to cause it to fragment into neutral and charged species. These issues are discussed in the paper entitled Regulatory Mass Spectrometry. Figure 11

How do you do trace organic analysis?

The sequence of events follows this order. Samples are collected in the field based on a rational sampling plan (Figure 12) resulting from some concern about human health or the environment. The samples are analyzed directly or are extracted into a solvent, cleaned up (removal of nontarget substances or interferences to analytes), and subjected to the final separation/detection technique. For example, a sampling plan that follows a plume of a known contaminant could be designed (e.g., pyrene), soil samples taken, extraction (e.g., Soxhlet or rapid extraction) and cleanup (e.g., GPC and silica get cleanup) into a final solution of methylene chloride, and determination by GC/MS using an appropriate internal standard and calibration curve (based on appropriate ions from the target analyte and internal standard). The apprehension of untargeted analytes follows a similar course but is more tenuous due to the potential of failing to recover compounds from unexpected chemistry behavior.

Figure 12

Solvent Properties and Polarities on Si

*Methanol, dioxane, and acetonitrile are insoluble in hexane.

$chloroform/methanol/acetone in the w/w/w of 47/23/30 corresponding to 31.50/29.08/37.93 v/v/v form an azeotrope that extracts compounds with a broad range of polarities.

Solvent Strengths on C18

Solvent Strengths on Carbon

Some links:

http://home.planet.nl/~skok/techniques/hplc/eluotropic_series_extended.html

Aldrich Chemical Company Catalog

Interpretation of Mass Spectra

Consider the mass spectrum of acetone given below.

Figure 13

First examine the molecular ion region (M+.) and the isotopes visible here (m/z 58). Usually, M+1 and M+2 are most important but in the case of halogenated compounds or organometallic compounds, for example, significant (in terms of relative abundance) isotope peaks occur at even higher m/z values. In the case of acetone the M+1 peak at m/z 59 is about 3-4 % relative abundance and the M+2 is 0.2-0.3% relative abundance to the M+. peak. The presence of Si or S is usually evident in the M+2 peak relative abundance. The presence of O or N is not obvious from relative abundances.

The first major loss of 15 corresponds to loss of CH3. and production of (CH3CO)+ at m/z 43. There is also an ion at m/z 42 corresponding to (CH2=C=O)+. known as ketene. The ion at m/z 15 is CH3+. Ions such as m/z 43 and 42 are common in mass spectra and as neutral losses as well but the identity may be hydrocarbon (CH3CH2CH2+) instead of an oxygen-containing moiety. This is a general problem in fragmentation studies of identifying the elemental composition of ions. Additional collisional study of ions or linked scans allows mapping of fragments to their parent ions. One final note before leaving this example: the absence of ions at m/z values is also important with regard to specificity.

Consider next the mass spectrum of the methyl ester of stearic acid. The molecular ion region supports the presence of 18-20 carbon atoms in the compound. The fragmentation produces two homologous series of ions. One set contains the carbonyl function (m/z 74 and m/z 87, 101, 115, 129, 143, 157, 171, 185, 199, and 213) and while the other contains pure hydrocarbon series (43, 57, 71, 85, 99, 113 etc and the same ions associated also with loss of H₂ such as m/z 55, 69, 83).

Figure 14

Interpretation of mass spectra then includes an examination of the molecular ion region, the consideration of rational cleavages and ion series, the recognition of particular m/z values that are correlated with commonly observed ions, and knowledge of the specific fragmentation pathways of compound classes and functional groups. This process is enormously aided by the existence of computer searchable libraries of spectra that now number more than 275,000 spectra. The hypothesis of a structure for a complete unknown is difficult and frequently benefits from the exact mass measurements of the molecular ion and key fragment ions. Even so, the identification may require purchase of standards for careful study, synthesis, or other spectroscopic data for a complete elucidation.

Ref and links:

F. M. McLafferty,

Analytical Problem Approach

1. Target analyte(s) or characterization study; what are the volatility, stability, and solubility of the analytes?

Look at the last step of the analysis. What determinative step is going to be used or is suggested by the properties of the analytes? Will derivatization be used?

2. Final separation/detection choice:

A. Target solubilities, volatility-stability. GC or GC/MS or derivatization and GC/MS. Ionization mode: EI, PICI, NICI (EC)

B. HPLC (detector) reverse phase or normal phase; LC/MS.

C. CE (detector) free zone, MEKC, or other; CE/MS.

D. TLC (detector) and solvent system for separation.

E. Other

3. Extraction and cleanup choices depend on matrix:

Water (liq/liq extraction, SPE, freeze dry)

Soil/sediment (Soxhlet, rapid extractions, sonication, shake outs)

Tissue (same as soil but additional considerations often having to do with biological substances such as lipids, carbohydrates, and proteins)

Other (e.g., monomers in a polymer)

Cleanup: normal phase (polar adsorbent) or reversed phase (hydrophobic adsorbent). Normal phase on Si uses a hexane, methylene chloride (or ether), acetone or methanol mixtures to elute. Reverse phase uses water and methanol or acetonitrile to elute. Adsorbents include silica, alumina, florisil, and carbon in normal phase and C18, C8, C2, and several others in reverse phase. GPC cleanup and HPLC cleanup may be used in multidimensional approaches.

Calculation of Isotope Abundances

The isotope abundances of a given element are those found for them in the earth's crust. For example, chlorine 35 and 37 have relative abundances of 0.7577 and 0.2423 for a total of 1.0 when added together. These percentages may vary depending on the location of a particular sample and whether biological processes have occurred that may alter the ratio. Thus, the accuracy of isotope ratio information is not comparable to the accuracy of the masses of ions. Nevertheless, much use can be made of this information. Some elements have, of course, only a single isotope (e.g., F and I). Other elements may have as many as 10 separate isotopes (e.g., Sn) and this can enormously complicate the spectra and the calculation of the isotope distribution.

The isotopic distribution is governed by the binomial distribution when calculating what the isotope pattern will be for a given set of elements (see http://www.mathsdirect.co.uk/pure/purtutbingen.htm for a discussion of the mathematical basis of the expansion). The general formula is:

Figure 15

For one chlorine the pattern is simple: 0.7577 / 0.2423 or 1.0 to 0.320 is the ratio of the relative abundances in the spectrum.

For two chlorines, we need to consider that we may have two ³⁵Cl₂, ³⁵Cl³⁷Cl, and ³⁷Cl₂ compositions possible. In order to calculate the distribution, we use the coefficients and powers dictated by the binomial theorem and the isotope abundances in the calculation:

Figure 16

(0.7577) x (0.7577) ; (0.7577) x (0.2423) ; (0.2423) x (0.2423) . We also need to count the mixed one twice for two different ways to arrange the chlorines. This is also the multiplier which falls out of the binomial theorem given above and begins to explain why this theorem describes the way isotopes are distributed. Normalized we get 1.00, 0.6396, and 0.1023 for the relative abundances of a two-chlorine cluster. This process can be carried out for multielement species with some complications of course.

It is interesting to note that for organic compounds, the carbon isotope is always a major consideration. When the carbon numbers start to get very high, the isotope distribution becomes dominated by the carbon content. For example, a 100 carbon molecule already has as base peak the M+1 ion as illustrated for C₁₀₀H₁₀₀:

Figure 17

For an online calculator, see http://www.shef.ac.uk/chemistry/chemputer/isotopes.html

Mass Spectrometry Data Acquisition, Processing, Quantitation, and Identification

Mass spectrometers, no matter their particular design (magnetic sector, quadrupole, ion trap, time of flight, etc.) for the purposes of this discussion have two acquisition modes: full or partial mass scans and ion monitoring (variously called multiple ion detection (MID), selected ion monitoring (SIM), selected ion recording (SIR), and single ion monitoring). SIM will be chosen as representing the latter mode of operation.

Data is acquired by digitizing the signal from the mass spectrometer at the rate of 40 kHz to 500 KHz, depending on sampling intervals dictated by mass resolution and scan speeds. Peak detection algorithms process the digital signal and result in centroided data that allow the assignment of a mass by means of a calibration that associates time with a given mass value (there could also be other mass assignment approaches such as field probe based assignments). Individual mass ions can be called out and displayed in a chromatographic format so that a particular ion current or m/z can be viewed as a function of time on the chromatographic time scale (i.e., retention time for ion chromatograms). These displays can be processed and areas of peaks obtained to use for quantitative determinations using peak detection algorithms. These algorithms usually take as parameters a peak detection sensitivity parameter for sensing peak start, a peak width parameter, and sometimes a peak end parameter for terminating the tail.

A calibration plot can be obtained in a number of ways. Illustrated below are some data for the calibration plot for a response of an ion from clobibric acid versus an ion of an internal standard. The data are handled by plotting the ratio of the response of the analyte to that of the internal standard versus the ratio of the amount of analyte to that of the internal standard. A linear regression of some order (usually linear, quadratic, or cubic) is performed on the data to obtain a calibration line.

Figure 18

y = 2 x + 0.001

y = (resp)s / (resp)is

(amt)s = (y * (amt)is - 0.001) / 2

In the case of the real sample we know the equation of the line, and all values except (amt)s. The value of (amt)is is 10 ppb for the example above and the resp values are the areas of the peaks from the ion chromatograms of target and is. Note that for quadratic and cubic fits the calculation is more complicated because a quadratic or cubic equation must be solved. A root finding numerical approach can be used rather than exact analytical expressions.

In the case of an identification that is based on the correspondence of the mass spectrum and retention time of a standard and a response in a sample, the retention times should be identical within reproducibility and the spectrum should also be quite similar. The specificity of the particular mass spectrum under consideration is also a critical element in the confirmation of identity.

Equations of Chromatography