Gas Chromatography-Mass Spectrometry

Gas Chromatographic-Mass Spectrometry Is Olive Oil the healthiest oil?

Introduction

Gas chromatographic-mass spectrometry (GC-MS) is a very powerful and ubiquitous analytical technique. It is often the analytical method of choice in toxicology, forensics, food science, and environmental research. In essence, this hybrid instrument replaces the traditional thermal-conductivity detector (TCD) or flame-ionization chromatographic detector (FID) with a very sensitive and information-rich mass spectrometer (MS). Not only can a GC-MS separate the volatile components of complex mixtures, but it can also record a mass spectrum of each component. This hybrid instrument provides two separate dimensions of information about the components in the sample, GC retention times and electron ionization (EI) mass spectra. GC retention time is related to specific chemical properties of the molecules in question (e.g. volatility, polarity, presence of specific functional groups) while molecular weight (derived from the mass spectrum) is indicative of atomic composition.

Chromatographic techniques separate mixtures of species based on their interactions with a stationary phase and a mobile phase. In the A315 gas chromatography experiment, volatile and semi-volatile species are vaporized into a flow of helium (mobile phase) and blown through a fused silica open tubular capillary (stationary phase). The capillary is usually derivatized to control its polarity (ranging from hydrophobic dimethylsiloxane to hydrophilic polyethylene glycol or any of a wide array of specialty coatings). The gas phase molecules interact with the surface and exist in dynamic equilibrium between being adsorbed onto the inner wall of the column and dissolved in the carrier gas. Gas chromatographs most often use temperature to control this equilibrium (higher temperatures shift the equilibrium more towards the mobile phase while lower temperatures shift it to the stationary phase). This is analogous to changing the polarity of the solvent mixture in thin-layer chromatography (TLC). Various components of mixtures will have different equilibria with the stationary phase and will consequently move through the column at different rates. The temperature can be left constant for maximum resolution or varied to minimize analysis time. A detector at the end of the column records the amount of time required for a given compound to elute off of the column (cf. RF values in TLC). With the proper conditions and column chemistries, very similar compounds can be separated by GC (e.g. isomers of dichlorobenzene, cis- and trans- fatty acids, racemic mixtures, etc.).

The GC-MS places an electron ionization quadrupole mass spectrometer at the end of the column. The mass spectrum provides two valuable pieces of information; the retention time and the EI mass spectrum. Molecules eluting from the column are directed into the source of the mass spectrometer and ionized by a beam of 70 eV electrons from a hot filament (M + e- M+• + 2e-). It is important to point out that EI generates odd-electron radical cations. The resulting positive ions are then analyzed by a quadrupole mass filter. Ideally, the molecular weight of the compound can be directly inferred from its mass spectrum. The mass spectrum indicates the mass to charge ratio of the ions, not the molecular weight of the neutral species (amu/e or Da/z

where e is the charge on an electron, Da is Daltons (1 Da = 1 amu) and z is the number of positive charges). However, the vast majority of ions generated by EI are singly charged and the charging is accomplished by removing an electron (an electron weighs 0.0048 amu), thus the masses of the ions in an EI mass spectrum are nearly identical to the neutral species from which they are derived. Because EI is a rather energetic process, most organic compounds fragment in the MS ion source, therefore the EI mass spectrum also indicates the masses of these fragments. While this may seem to be a weakness at first (spectra are more complex and sensitivity for the intact species will suffer), a careful analysis of a fragmentation pattern can be used to determine the connectivity of the atoms in the original molecule. In addition, the GC-MS system in the 315 laboratory contains a computerized library that can match mass spectral “fingerprints” with known spectra to aid in the identification of unknowns. Figure 1 contains the mass spectrum of ∆9- tetrahydrocannabinol (THC, the hallucinogenic agent in marijuana) from the NIST ’02 mass spectrum library. The NIST ’02 database contains the 70 eV EI mass spectra of over 150,000 compounds.

The Quadrupole Mass Filter and Spectral Skew The mass analyzer of the Agilent 5973 inert mass-selective detector is a quadrupole mass

filter. As the name implies, this device allows only a narrow range of ions (having a specified mass-to-charge ratio, m/z) to pass through at any given time. Thus, the quadrupole is directly analogous to the grating monochromator used in optical spectroscopy techniques. Consequently, quadrupole mass spectrometers are scanning instruments. To generate a complete mass spectrum, they pass ions of a particular m/z for a brief period of time, record the ion current, move on to the next m/z value, and repeat the entire process over the specified range. Depending on the range of m/z ratios scanned, the electronics driving the instrument, and the quality of data required, a quadrupole mass analyzer may require 0.1–10 sec to construct a single mass spectrum (i.e., spectra are created at a rate of 0.1–10 Hz).

When a quadrupole mass analyzer is used to monitor the output of a chromatographic column (as in GC-MS), a phenomenon referred to as spectral skew results. As compounds elute from the column, they are physically separated into bands which arrive at the detector in

(mainlib) Dronabinol20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320

0

50

100

29

41

55 67 81 91 107 121 147 160 174 193 217

231

243 258

271

299

314

OH

O

Figure 1: 70 eV mass spectrum of delta-9 tetrahydrocannabinol

sequence. As viewed from a fixed point near the end of the column, the concentration of any given compound will appear to rapidly rise and fall upon the arrival of its band, creating the familiar chromatographic peak. Since a quadrupole mass analyzer requires a finite amount of time to monitor each m/z ratio, the intensities of different ions will be recorded at different points within a single chromatographic peak. Depending on which direction the quadrupole is scanning and where the scan lies within the chromatographic peak, the intensities of either larger or smaller m/z ratios are artificially inflated relative to other ions in the mass spectrum, causing a skewed spectrum to be recorded. The critical question one must ask to assess spectral skew is, “How does the scan rate compare with the width of a typical chromatographic peak?” If the scan time is not significantly less than the width of a peak, spectral skew will be problematic. One way to reduce the effect of spectral skew is simply to increase the scanning speed of the mass analyzer. For a quadrupole mass filter, this is most often achieved by narrowing the range of m/z ratios scanned. As the time required to obtain a single scan becomes smaller, the width of the chromatographic peak traversed during the scan becomes smaller, and the degree of skew is lessened. If mass spectra can be recorded instantaneously (which can be approached by some time-of-flight [TOF] mass spectrometers), spectral skew is eliminated entirely.

SCAN and Select-Ion Monitoring (SIM)—The 5973N GC–MS quadrupole mass analyzer can be operated in two modes: SCAN and select-ion monitoring (SIM). In the SCAN mode, the quadrupole continuously and repeatedly ramps the monitored m/z ratio from a preset lower limit to a preset upper limit, generating a series of complete mass spectra. At the conclusion of each individual scan, the intensities of all the m/z ratios within the scan are summed, giving a total ion current. A chromatogram is then constructed by plotting the series of total ion current versus retention time. This plot is called the total ion chromatogram, or TIC. Analyzing a GC–MS chromatogram obtained in the SCAN mode therefore consists of selecting the portion of the TIC that corresponds to a given peak and extracting the mass spectra from that time period.

In the select-ion monitoring (SIM) mode, the quadrupole remains fixed on a small set of m/z ratios, effectively allowing only those predetermined masses to pass through to the detector. An analysis in the SIM mode is useful when one is looking for small quantities of known compounds under circumstances in which they cannot be separated from other compounds chromatographically (i.e., in very complicated mixtures). In addition, the SIM detection scheme often yields substantially lower detection limits than the SCAN mode (more time is spent monitoring the m/z of interest) and also minimizes spectral skew.

Automated Data Collection—The 6890 GC is fitted with a 100-position autosampler which allows the GC-MS to record data without user intervention. Analytical chemists use autosamplers to increase the number of samples run by an instrument over the course of a day (injections can be made after the technicians have gone home) and to increase productivity (analysts can perform other tasks while the autosampler is working). Furthermore, they free the analyst to focus on data interpretation rather than data collection. Another advantage of robotic sample handling is that it is ensures that all injections are made in a highly reproducible manner. Human users may inject 1.5 uL or 0.95 µL instead of 1.0 µL or push the plunger with a different

speed for each injection. These small variations can lower the precision of the data. The robotic injector always injects samples the same way every time. Another useful feature is that autosamplers can perform many different analyses within a sample set. For example, one can direct the instrument to record data using a SCAN mode method followed by a second analysis in SIM mode method. As long as a hardware change (e.g. switching column chemistries) is not required, the autosampler can perform an analysis. You will take full advantage of these capabilities. The instrument will continue to inject samples after the assigned laboratory period is over. The data can be analyzed at any time on any computer that has Agilent’s ChemStation 1701DA software. Analysis of Mass-Spectral Data - Important terms and ideas:

1. Molecular ion peak (M+) or parent ion peak: the peak that results from the loss of an electron from the molecule. Since the electron is of negligible mass, the M+ peak gives you the mass of the original molecule. Note that you may not be able to find a molecular ion peak in mass spectra of molecules that fragment easily (e.g., aliphatic hydrocarbons). Furthermore, the M+ peak is considered the mass peak containing atoms of the most abundant isotope for each element (i.e., 12C, 1H, 16O, 35Cl, 79Br, etc.) and not the less abundant isotopes (i.e., 13C, 37Cl, 81Br, etc.). This is often one of the highest m/z ratio ions in an EI mass spectrum. The official name for this peak is the “monoisotopic peak.” The second peak in an isotope cluster is called the M+1 peak; the third peak in an isotopic envelope is called the M+2 peak, and so on. Be aware that co-eluting compounds, background contamination in the ion source, and portions of the column stationary phase bleeding into the mass spectrometer can create high m/z ions.

2. Base Peak: the largest peak (highest abundance) in the spectrum. The abundance of each

mass peak in the spectrum is representative of the ion’s stability in the source, thus the base peak is almost always the most stable ion. Occasionally, the base peak is the same as the molecular ion peak, but this not always the case. Since aliphatic compounds tend to fragment extensively, the base peak is usually a fragment of the original compound.

3. Recall that the atomic weights reported on the periodic table are weighted averages for all

isotopes. The mass spectrometer can distinguish between the isotopes. Thus, use the mass of the predominant isotope for each atom. For example, the mass of 12C (the most abundant isotope of carbon with a nominal mass of 12) is 12.0000 (not 12.0107) Da (Daltons, 1 Da = 1 atomic mass unit), the mass of 1H is 1.0078 and the mass of 16O is 15.9949 (the resolution of the 5973 inert is such that you can round off to the nearest 0.1 Dalton). These differences may seem slight, but they add up quickly in a large molecule (e.g. for octane, C8H18 has a monoisotopic mass of 114.1409, not 114.0000 (0.1409 = 0.007825 * 18)). The molecular weight calculated using mass values from the periodic

table is called the “average mass.” This is the molecular weight written on the sides of reagent bottles and used for stoichiometric calculations.

4. The intensity of a particular ion is proportional to its concentration in the source at any

given time. Thus, a doubling in the intensity of a particular ion implies that the concentration of the molecular species responsible for that ion has also doubled. This is analogous to Beer’s Law used in absorption spectroscopy and enables the analyst to use the same quantification methodologies used for GC-FID or GC-TCD experiments for GC-MS. You will calculate response ratios for various fatty acids to allow you to compare their abundances in commercial cooking oils.

Steps to identifying a compound based on its mass spectrum:

1) Preliminary survey a. How many carbon atoms are present? b. Are there any N, Cl, Br, or S atoms? c. Is the molecule aromatic or aliphatic? d. What is the molecular weight of the molecule?

2) What functional groups are present? 3) The molecular formula can be determined by piecing the functional groups together.

Hydrogen and oxygen can be filled in as needed. 1a. How many carbon atoms are present?

Recall that most carbon is in the form of 12C (12.0000 amu), but 1.1% of carbon is 13C (13.0034 amu). Start with the M+ peak. This is the monoisotopic mass of the molecule. In carbon containing molecules there will be a less abundant M+1 peak due to the presence of 13C (~1 Dalton heavier). The rule of thumb that follows does begin to break down for molecules larger than 500 Da (as the chance of a molecule having two or more 13C’s becomes significant), but very few compounds with molecular weights greater than 500 are volatile enough for gas chromatography. Although the plotted mass spectrum is useful for obtaining an overview of the main spectral features, the tabulated list of mass-to-charge ratios (m/z) versus abundance values is essential for an accurate determination of the number of carbon atoms. Below is an example of how the number of carbons in a given molecule can be calculated:

Peak: m/z: Abundance: Abundance as a % of M+ M+ 90 73.0 73.0/73.0*100 = 100 M+1 91 3.3 3.3/73.0*100 = 4.5

Number of carbon atoms = relative abundance of M+1 peak / 1.1

Using the numbers from above: # of carbons = 4.5/1.1 = 4 carbons. 1b. Are there any N, Cl, Br or S atoms? According to the “nitrogen rule,” a molecule with an odd M+ must contain an odd number of nitrogen atoms and an even M+ implies an even number of nitrogen atoms (0 counts as an even number in this case). For Cl, S, and Br, look for at the M+2 peak. These elements, like carbon, have heavy isotopes that can be detected by the mass spectrometer. However the heavy isotopes of these atoms are two Daltons heavier than the most abundant isotope. Lack of an M+2 peak is a clear indication that the molecule does not contain these atoms. If there is a significant M+2 peak present, then you need to look a little closer at the ratio. If (M+/M+2 = 3), then a single chlorine is present; if (M+/M+2 = 19), then one sulfur is present; if (M+/M+2 = 1), then one bromine is present. These ratios are based on the relative abundances (just as in the case of carbon) of each heavy isotope. While these ratios are useful, there is one note of warning that ought to be made about using them to identify which atom is present. If there are multiple Cl, Br or S atoms present in the molecule, then a distribution of peaks in a series (M+, M+2, M+4, M+6, etc.) will result. In these cases, the intensities of these peaks will follow a statistical distribution that is more complicated than the simple ratios above imply. However, in this case there are other ways to determine which atom is present. Look for losses of 35 (for chlorine) or 79 (for bromine) since halogens usually fragment from molecules rather easily.

1c. Is the molecule aromatic or aliphatic?

Since aromatic compounds tend to be quite stable, they show less fragmentation under electron impact ionization. Thus, you will see more intense peaks at higher masses with a strong M+ peak. If the compound is primarily a substituted benzene ring, then a strong m/z = 77 is certain to show up in the spectrum due to the C6H5

+ fragment ion. An aliphatic compound will show extensive fragmentation and exhibit a very weak M+ peak. Straight chain aliphatic compounds should show a series of peaks at m/z = 15 + 14n, where n is the number of CH2 groups in the chain and m/z = 15 corresponds to the terminal methyl group. 1d. What is the molecular weight?

By now it should be clear that the molecular weight of the molecule corresponds to the m/z of its M+ peak. 2. What functional groups are present?

Frequently the functional groups on the molecule can be determined by searching for particular patterns. Here are some examples:

a) Alcohols: M = 31 + 14n (formation of R-CH=OH+, where R is composed of n CH2 groups)

b) Aliphatics: M = 15 + 14n c) Benzene: M = 78 (subtract one mass unit for each substitution on the ring)

d) Ketones: M = 43 + 14n (formation of R-CH2-C≡O+)

More examples can be found in textbooks (e.g. Interpretation of Mass Spectra, 4th edition by F. W. McLafferty and F. Tureček). Sometimes identification is not this straightforward. It is not uncommon for a hydrogen atom or a methyl group to migrate to form a more stable ion. Generally, fragmentation will occur where functional groups connect, such as places where branching occurs or where a side chain is attached to an aromatic ring. 3. Piecing it all together…

At this point you ought to have some idea of what your molecule looked like before fragmentation. You should know the number of carbon atoms and whether or not any N, Cl, Br or S atoms are present. Any mass left over can be accounted for by adding oxygen (M = 16) and hydrogen (M = 1) until you manage to assemble a coherent structure for the molecule. Figure 2 is an example.

The first thing to notice is that the base peak is M = 94. The M+ peak is M = 136.

Figure 2: Mass spectrum of unknown 136 Da compound

Based on the data the molecular weight of the compound is 136. Computing the number of carbons yields:

2.5/25*100 = 10 = M+1 relative to M+ peak Therefore, M+1/1.1 = 10/1.1 = 9.09 ≈ 9 carbons There is no significant M+2 peak, so there likely aren’t any chlorine, bromine or sulfur atoms in the molecule. Since M+ is even, there is either an even number of nitrogen atoms or none at all. Notice that there is an intense m/z = 77 peak, which corresponds to a mono-substituted benzene. This leaves three other carbons to figure out. Molecular weight = 136 -77 from the benzene ring -36 (from 3 12C atoms) ____________________ 23 mass units left over There must be an even number of nitrogen atoms present, however since two nitrogen atoms (m/z = 14) would give a mass of 28, the possibility of nitrogen can be eliminated. The remaining mass can be accounted for by filling in with oxygen and hydrogen. Obviously the three remaining carbons are not going to carry 23 hydrogen atoms, however if we use one oxygen atom: 23 (mass unaccounted for) – 16 (oxygen) = 7 mass units left over. It is reasonable to imagine seven hydrogen atoms on three carbon atoms. With all of the mass accounted for the next step is to propose a likely structure. There are several possible combinations, but only a few actually make sense. Perhaps the molecule is an alcohol. Recall that alcohols show a 31 + 14n mass pattern. There are no m/z = 31, m/z = 45, or m/z -= 59 peaks in the spectrum, so it is probably not an alcohol. How about a ketone? There is an m/z = 43 peak present. However if the molecule were a ketone then it could not accommodate seven hydrogen atoms. What about an ether? That seems to work. What if the structure were: (C6H5)-CH2-O-CH2-CH3?

As mentioned above, molecules generally fragment between functionalities. If this were the actual structure, we would probably see an intense fragment with M = 91, resulting from the bond breakage between the leftmost CH2 and the oxygen atom. However there is no fragment at this mass in the spectrum. Perhaps the molecule is an isomer: (C6H5)-O-CH2-CH2-CH3

If the molecule fragmented such that the oxygen was attached to the aromatic ring, but the rest of the chain broke off, then these fragments would be left:

(C6H5)-O• and +CH2-CH2-CH3.

Generally a hydrogen atom will shift to stabilize an oxyanion, resulting in this fragment:

(C6H5)-OH+ (m/z = 94) Referring back to the mass spectrum, m/z = 94 is the most prominent peak, making this structure the most likely candidate. Thus the final proposed structure is (C6H5)-O-CH2-CH2-CH3.

Clearly, the interpretation of mass spectra takes a lot of finesse. If you were to do this more frequently, common fragmentation patterns would become much more obvious to you. Fortunately modern chemists can utilize the on-line spectral libraries of GC-MS systems to narrow down the likely choices. The Mass Spectrometry Facility has purchased a copy of the NIST ’02 library which contains reference spectra for over 150,000 compounds. Most database searching programs arrive at their suggestions by comparing the analyte mass spectrum to all of those in the library. However it is essential to understand how to properly interpret mass spectra in order to properly assess the validity of a computer-generated database matches.

4. Incomplete Mass Spectra Although mass spectra databases are very powerful tools for identifying the components of a complex mixture, they are far from perfect. No database is complete, and there will be times (especially when synthesizing novel compounds) that the molecule of interest simply will not be in the database. The NIST ’02 library contains over 170,000 entries (150,000 compounds, there is some redundancy in this database), but there are millions of potential organic compounds. Another challenge arises when a component of a mixture is present at very low levels. Often the mass spectra observed will be incomplete. That is, only the most intense fragment ions in the mass spectrum will show up because the other peaks will be too weak to be considered by the software. One solution to this would be to inject a more concentrated sample, thus increasing the

concentration of the component in question. This is not always an option however as analysts are often sample limited in the real world. Furthermore, injection of a more concentrated sample may foul the column or EI source of the GC-MS significantly shortening the lifetime of those components. Remember, the concentration of the major components in a mixture will increase as well. Figure 3 contains the mass spectra of dodecane, 1-dodecene, and 1-dodecyne and an unknown, low abundance compound with 14 carbons. Look at the three 12-carbon molecule spectra and notice the differences in the masses of the low mass (<100) fragment ions.

By comparing the low mass portion of the unknown spectrum to the other three, the identity should be clear. Check with your AI during the lab period for the correct answer. A similar analysis can be to identify compounds with shared structure features. In this experiment, one of the components of olive oil will likely be insufficiently abundant for a proper database match. You will have to compare its mass spectrum to the mass spectra from olive oil’s other components to determine its identity.

Response Factors and Quantitation As stated above, the response observed for a given compound is proportional to its concentration in the gas stream. Unfortunately, not all compounds are ionizable or detectable with similar efficiencies. This problem is also encountered in optical spectroscopy. Recall that Beer’s law contains a proportionality constant (the molar absorptivity, ε) that relates how well any compound absorbs the wavelength in question. If ε is not readily known, a line relating peak area and concentration of a compound can be created by analyzing a handful of standards with known concentrations of the compound in question. This method can become cumbersome as the number of analytes in a given sample increases. An alternative to generating a calibration curve for each analyte is computing response factors. A response factor is a ratio between the observed response (peak area) for a given concentration of a molecule with the response from the same amount of a closely related standard (e.g. a stable isotope labeled version or a structural isomer). According to method 8270C from the United States Environmental Protection Agency, the response factor is computed as follows:

The response factor is used to normalize the peak areas of the other analytes using this equation: In this experiment, you will normalize all response factors to that of palmitic acid methyl ester. The corrected areas will then be used to compute the compositions of the oils. Fatty Acid Analysis

The calorie, vitamin, mineral, protein, carbohydrate, and fat content of nearly every food item sold in the United States must be clearly shown on the package to allow consumers to make informed nutrition choices. These values are determined using a wide array of analytical chemistry techniques including atomic absorption spectroscopy, bomb calorimetry, gas chromatography, gravimetric analysis, liquid chromatography, and ultraviolet-visible

Figure 3: Mass spectra of dodecane, 1-dodecene, 1-dodecyne and an unknown 14-carbon hydrocarbon. Note: the bottom 25% of the 14-carbon mass spectrum has been deleted to simulate the mass spectrum from a low abundance compound.

analyte

std

std

analyte

concconc

areaarea

RF ×=

analyte

analytecorrected RF

areaarea =

spectroscopy. This particular experiment will use GC-MS to determine the fatty acid profiles of a handful of commercially available cooking oils.

For most foods, a chemical method is required to isolate the nutrient of interest from the food prior to analysis (typically homogenization followed by refluxing solvent extraction (aka Soxhlet extraction), supercritical fluid extraction, solid phase extraction, enhanced microwave extraction, etc.). The extracted nutrients are often then further modified to make them more amenable to a particular analysis method (e.g. a chromophore may be added, the nutrient may be converted to an easily quantified small molecule, the nutrient may be esterified, etc.). Fats exist as triglycerides, three fatty acids attached to a glycerol molecule with ester bonds. Fatty acids are long chain carboxylic acids that are isolated from naturally occurring fats. They can be saturated (fully hydrogenated, i.e. all carbon-carbon bonds are single bonds), mono-unsaturated (one double bond between carbon bonds in the chain), and poly-unsaturated (more than one double bond in the carbon chain). All double bonds in natural fatty acids are cis isomers. The degree of saturation determines many important properties of the fat including flavor, resistance to oxidation, melting point, and smoke point. Olive oil, which contains more than 70% unsaturated fatty acids melts at -6 °C and has a smoke point of about 210 °C. Refined coconut oil (8% unsaturated fatty acids) melts at 25 °C and does not smoke until 232 °C. Food processors often partially hydrogenate vegetable oils to raise their melting points, lengthen their shelf lives, or change the texture of the finished product. The high temperatures and pressures of the hydrogenation process can cause isomerization in unsaturated fatty acids, creating trans fatty acids. Table 1 contains a list of common fatty acids and their sources.

There is an overwhelming body of evidence concerning the health effects of the various types of fatty acids, and as a result food labels must now list the amount of saturated, unsaturated, and trans fatty acids in each serving. These effects are summarized in an American Heart Association publication (http://www.webmd.com/content/article/124/115606). Saturated fatty acids have been shown to raise low density lipoproteins (LDL or “bad” cholesterol). Cis isomers of mono- and poly-unsaturated fatty acids have been shown to lower blood cholesterol levels. Trans isomers of unsaturated fatty not only raise LDL but lower heart protecting high density lipoproteins (HDLs or “good” cholesterol); some researchers consider them worse than saturated fats. The federal government has mandated that the saturated, unsaturated, and trans-fat content of all foods be included on the label. It also stipulated that foods with less than 0.5 g of trans fatty acids per serving could be labeled as having 0 g trans fat.

Table 1: Names, properties, and sources of common fatty acids.

Common Name Carbon atoms

double bonds IUPAC name Sources

Butyric acid 4 0 butanoic acid Butterfat Caproic Acid 6 0 hexanoic acid Butterfat Caprylic Acid 8 0 octanoic acid coconut oil

Capric Acid 10 0 decanoic acid coconut oil Lauric Acid 12 0 dodecanoic acid coconut oil

Myristic Acid 14 0 tetradecanoic acid palm kernel oil Palmitic Acid 16 0 hexadecanoic acid palm oil

Palmitoleic Acid 16 1 9-hexadecenoic acid animal fats Stearic Acid 18 0 octadecanoic acid animal fats

Oleic Acid 18 1 cis 9-octadecenoic acid olive oil Elaidic Acid 18 1 trans 9-octadecenoic acid Hydrogenated oils

Ricinoleic acid 18 1 12-hydroxy-9-octadecenoic acid castor oil

Vaccenic Acid 18 1 11-octadecenoic acid Butterfat Linoleic Acid 18 2 9,12-octadecadienoic acid grape seed oil

Alpha-Linolenic Acid 18 3

9,12,15-octadecatrienoic acid

Flaxseed (linseed) oil

Gamma-Linolenic Acid 18 3 6,9,12-octadecatrienoic acid borage oil

Arachidic Acid 20 0 eicosanoic acid peanut oil, fish oil Gadoleic Acid 20 1 9-eicosenoic acid fish oil

Arachidonic Acid 20 4 5,8,11,14-eicosatetraenoic acid liver fats

EPA 20 5 5,8,11,14,17-eicosapentaenoic acid fish oil

Behenic acid 22 0 docosanoic acid rapeseed oil Erucic acid 22 1 13-docosenoic acid rapeseed oil

DHA 22 6 4,7,10,13,16,19-docosahexaenoic acid fish oil

Lignoceric acid 24 0 tetracosanoic acid small amounts in most fats

Table adapted from http://www.scientificpsychic.com/fitness/fattyacids.html

In this experiment, we will saponify the oils to release the fatty acids. Saponification, or

treating the fat with a strong base, hydrolyzes the triglyceride, creating one molecule of glycerol and three fatty acid molecules. Since the free fatty acids are not very volatile, they will be converted into methyl esters by reacting them with boron triflouride in methanol. The resulting fatty acid methyl esters (FAMEs) are then injected into the GC-MS. Mass spectra will be used to determine the identities of the FAMEs; GC retention times will be used to discriminate between various FAME isomers.

Gas Chromatography-Mass Spectrometry Experiment

Materials

Glassware and other equipment: 5 8” test tubes 1 250 mL beaker 1 hot plate 1 10 mL graduated cylinder 2 25 mL graduated cylinders 10 125 mL Erlenmeyer flasks with stoppers 100 uL variable pipette 1 mL pipette 7 autosampler vials

Reagents: 75 µL of olive oil 75 mg of 0 g trans fat shortening 75 µL/mg of 3 different fats of your choosing (not olive or 0 g trans fat shortening) 18 mL of 0.5 M NaOH in CH3OH 24 mL of 14% BF3 in CH3OH 120 mL of saturated NaCl in H2O 150 mL of hexane 100 g of anhydrous Na2SO4 1 mL of standard FAME mix in hexane 1 mL of 25 ppm oleic acid (cis 18:1) methyl ester in hexane 1 mL of 25 ppm elaidic acid (trans 18:1) methyl ester in hexane

Procedure

(adapted from J. F. Robinson and J. Neyer-Hilvert in J. of Chem. Educ. 74:1106-1108 (1997))

Part 1: Saponification of the Oils

1) Place a half-filled 250 mL beaker of water onto a hot plate. Bring this to a slow boil. 2) Get small samples of olive oil, 0 g trans fat shortening, and three other fats for analysis.

NOTE: The solids should be scooped out with a spatula and put on weighing paper. You will need ~75 mg of the solid greases. Be sure to place the glob of grease at the BOTTOM of the test tube so it can react with the NaOH. Also, you will be unable to quantitatively transfer the solid greases; this is acceptable as all measurements in this experiment are relative to the total amount in the test tube. You will need 75 µL of each of the liquid fats.

3) Be sure to CLEARLY LABEL all test tubes, flasks and autosampler vials as all solutions in this experiment are clear.

4) Pour 3 mL of the NaOH solution into five 8” test tubes. 5) Add 75 µL/mg of each fat into a clean test tube and place it in a boiling water bath. 6) Boil them until the solutions are homogenous. 7) Remove the test tubes from the bath and let them cool for 3-5 minutes.

Part 2: Esterifying the free fatty acids 8) Add a boiling chip and 5 mL of 14% BF3/MeOH solution to each tube. 9) Boil the solution for 5 minutes in the water bath. 10) Remove the test tubes from the bath and let them cool.

Part 3: Extracting the FAMEs from the reaction mixture 11) Pour 20 mL of saturated NaCl and 25 mL of hexane into five 125 mL flasks. 12) Pour each cooled reaction mixture into a separate 125 mL flask containing the NaCl

solution and hexane. 13) Seal the flask and shake vigorously for 2 minutes. You want to make sure the 2 layers

are mixed very well. Make sure to vent the flask occasionally to release the pressure. 14) Allow the layers to settle for ~1 minute. 15) Decant the hexane layer (the upper layer) into another 125 mL flask containing several

grams of sodium sulfate. 16) Add 25 mL of hexane to the salt/methanol solution and extract any remaining FAMEs

from the salt water and methanol. 17) Combine the 2nd hexane extract with the first extract.

Part 4: Diluting the FAME extract for GC-MS analysis 18) Add 20 µL of each FAME extract into separate autosampler vials containing 1 mL of

hexane and mix well. 19) Place 1 mL of each standard solution into separate autosampler vials. 20) Take all the autosampler vials and some clean hexane up to the mass spectrometry lab

(A411) for analysis.

Part 5: Recording Data with the GC-MS

21) Setting up the instrument for the experiments: If the software is not already open, log on to the computer using your network ID and password and start the ChemStation software by opening Instrument 1 (located on the desktop as well as in the Start menu). Select Method >> Load… and 315_FAME.m (it is in the directory “C:\MSDCHEM\1\315\Methods”). This step MUST be completed first.

22) Loading the autosampler: Place the autosampler vials containing the three standard

solutions into positions 1, 2, and 3 of the autosampler tray; put the five vegetable oil FAME extracts into positions 4-8; put an autosampler vial containing clean hexane into

position 9. Make sure the vial labeled “A” on the turret is full (nearly to the top) with clean hexane.

23) Creating the autosampler sequence table: This step creates the table that determines types

of analyses to perform on each sample and specifies the order in which they are to be performed. The table is generated as follows:

a. Click the “Open Sequence: icon (it has a folder and three autosampler vials in it)

and select the sequence BLANK.S from the C:\MSDCHEM\1\315\Sequences” directory.

b. Click on the “Edit Sequence” icon (it has a pencil and three autosampler vials). Make sure the data directory is “C:\MSDCHEM\1\315\Data\” and the method directory is set to “C”\MSDCHEM\1\315\Methods”.

c. Select sample type “Sample” for all entries. The other sample types are only used with automated data analyses. Enter “1” for the vial. Put a descriptive sample name in the “Sample” column (e.g. “standard FAME mix” or “corn oil extract”). Select the method 315_FAME after clicking inside the box under “Method/Keyword”. Enter a file name for the run. It is suggested that you use your initials and a three digit number (e.g. ABC_001). Enter a descriptive comment (similar to the “Sample” column) in the “Comment/Keyword String”. Leave the other columns (Sample Amt, Multiplier, Level, etc.) blank. When you are finished, your table row should look something like this:

d. Make a two more entries so the contents of the autosampler vial 1 are analyzed

three times (for computing averages and standard deviations). Make sure to increment the number in the file name. Create similar triplicate entries for each vial. You should analyze the blank only once with 315_FAME.

e. Your completed sample table should have 25 entries and should look similar to the table below.

24) Validate the Autosampler Sequence Table: This step ensures that no errors were made in typing the names of the methods and that there is enough room on the hard drive to record your data. Save your completed table by clicking the “Save Sequence” icon (it has a red disk and 3 vials on it). Save the sequence in the directory “C:\MSDChem\1\315\Sequences”. Give it a file name involving your initials so it will be easy to find again, if necessary. Click the “Simulate Sequence” icon (it has a red check mark and 3 vials on it). Make sure the “Full Method” button is selected in the “Method Sections To Run” box. Then click “Run Sequence”. Correct any errors that arise and resave your sequence. Make sure to view and print the verification report so you can check your notebook entries. Close the sequence verification report after you have printed it.

25) Start the Instrument: Click on the Run Sequence icon (it has a running figure and 3

autosampler vials on it). Make sure the “Full Method” button is selected in the “Method Sections To Run” box. Then click “Run Sequence”. The instrument will perform the 26 analyses specified in the sequence table.

At the end of each day

Dispose of any unneeded hexane solutions in the ORGANIC WASTE container. Dispose of the BF3/methanol/salt water solution in the AQUEOUS WASTE container. If you need to keep a hexane solution overnight, cap it and seal the top with a small strip of Parafilm. Make sure any flasks you keep are labeled and stored upright in the refrigerator or freezer rather than a drawer. Always clean your glassware, the balance area, and your bench space before you leave.

Part 6: Analyzing the GC-MS Data (Note: this section may be performed either in the Mass Spectrometry Facility, A454. See the Appendix entitled “How to use ChemStation in the Library for C315 Students” for instructions on remote access to the GC-MS computer. It also contains a brief introduction to ChemStation.)

26) Load the software and method: Open Enhanced Data Analysis by double clicking on its

icon. Load the 315_FAME.M method. 27) Analyze the Data from the FAME mixture:

a. Open the data from the standard FAME mixture. b. Click the “integrate” button on the software (it looks like a chromatogram with 2

peaks and an arrow pointing between them). If you are unsure what a particular button does, mouse over it for a few seconds, and a brief description of that button’s function will appear.

c. Print the chromatogram by clicking the printer icon. You will need this for your report.

d. Select “Integration Results…” from the Chromatogram menu. Copy the table by clicking the “Copy” button and paste the data into a spreadsheet for later processing.

e. Zoom in on each peak and extract its mass spectrum by dragging across the peak with the right mouse button.

f. Make sure the NIST ’02 library is selected and search the mass spectra against the NIST ’02 library. Use the results from the searches to add names to the table in your lab report.

g. Using the peak areas from the table and the equations in the handout, determine a response factor for each fatty acid. Use palmitic acid methyl ester as the basis of your factors. Ask your AI for the Certificate of Analysis for the FAME standard.

28) Determine the elution order for cis and trans 9-octadecenoic acid methyl esters: Open the

data file from the oleic acid methyl ester (cis 9-octadecenoic acid) solution. Note the retention time. Do the same for the elaidic acid methyl ester (trans 9-octadecenoic acid).

29) Determine the response factor for elaidic acid methyl ester: Using a procedure similar to that in step 26, compute the response factor for elaidic acid methyl ester.

30) Analyze the data from the oil samples:

a. Open the data file from the olive oil sample and integrate the chromatogram. b. Using both mass spectra and retention times, identify the components of each oil.

You will need to tabulate these data for your report. c. Using the response factors calculated in steps 26 and 27, compute the fatty acid

composition for each oil. Add these results to your table. Be sure to list all fatty acids more than 0.5% abundant.

d. Repeat steps 29a-29c for the remaining oil samples. e. For any fatty acid for which a response factor was not computed, assume a

response factor of 1.00 f. For the olive oil sample, determine the identity of the low intensity feature that

elutes just before palmitic acid methyl ester.

Part 7: Lab Report

1. Attach at the end of your report one representative chromatogram for each standard sample analyzed (standard FAME mixture chromatogram, oleic acid chromatogram, and elaidic acid chromatogram). DO NOT print each chromatogram – it is not necessary and wastes paper.

2. Create a table summarizing the chromatographic conditions used for the analyses. Include

the type of column used, the dimensions of the column, the injection volumes, split ratios, and the temperature program. Column data are posted on the instrument and the temperature program can be viewed from the Oven icon in the Acquisition window. Discuss how the method should be modified if the fatty acids had longer hydrocarbon chains.

3. Create a table summarizing the retention times for the FAMEs in the standard mixture.

Include the averages and standard deviations for the retention times for each FAME in the standard. In separate columns in this table, include the name, concentration (from the certificate of analysis), peak area (average and standard deviation), and response factor for each standard FAME.

4. Include in your report the mass spectra for the oleic acid and elaidic acid FAMEs. Describe

how you can determine which of the two isomers is present when only one is found in an unknown sample.

5. For each of the unknown oils, create a table that lists each of the fatty acid components with

a relative abundance of 0.5% or higher. This table should include name, retention time, peak area, corrected peak area, and relative concentration (average and standard deviation). Find a reliable reference and report the amount of each fatty acid “known” to be present in each oil. Compare the relative concentration you measured to this known amount. Comment on how well these numbers agree with each other.

6. Based on the data collected for your unknowns, which oil would you report to be the

healthiest? 7. In the chromatogram for olive oil, you should find a small peak with a retention time of

approximately 7.9 minutes. Include a mass spectrum of this compound in your report and any evidence that you can use to help identify this compound. Explain the quality of the mass spectrum and the impact of this quality on the data interpretation.

8. Based on your data table for trans-fat free Crisco, compute the number of grams of trans fat

in one serving (one tablespoon, which is 14 g for liquid oils and 12 g for solid oils/greases). Is this product truly trans-fat free? Comment on your findings.

9. Other than the balance and the glassware used, what are the major sources of error in GC-MS? How can these errors be minimized or eliminated? Does the autosampler address any of these issues? If so, how?

Gas Chromatography–Mass Spectrometry Report Grade Sheet

Section Points Qualifier 25 Introduction 10 Experimental 10 Results and Discussion

FAME Analysis (a) Chromatographic conditions

Table of chromatographic conditions 5 Conditions for longer chain fatty acid 5

(b) Standard results Chromatograms 5 Table of results 5 Response factors 5 Oleic & elaidic acid MS 5

(c) Unknown oils Table of results 10 Comparison to known values 10 Healthiest oil 5 Olive oil peak 5 Trans-fat free Crisco 5

(d) Sources of error 5 Conclusions 5 Literature references 5

Appendix: How to use ChemStation in the Library for C315 Students: ChemStation is installed on PC # 5 in the library. This PC is located next to the emergency exit at the far end of the library (near where the journal, “Tetrahedron Letters” is shelved). Login to PC #5 using your IU username and password. When the PC boots up, double click on the “Instrument #1 Data Analysis” icon. It should be on the desktop. If not, it can be found on the start menu in Programs Dept. Sponsored Chemistry MSD Chem Instrument #1 Data Analysis You will need to map C315 directory on the GC-MS computer in the mass spectrometry lab as follows: Click on the start bar and select Run… Put \\bl-chem-msfgcmf into the Open… dialog box. Right click on the “C315_Data_Methods” icon. Select Map Network Drive… Select a drive letter for the connection (e.g. W:). DO NOT SELECT “Reconnect at logon”! Click on Finish (see below)

Your data will now be on the drive with the letter you selected. For example, if you chose W:, your data will be found in the directory W:\Data. A copy of the NIST’02 library is also found in the C315 directory. Be sure to go to the “Spectrum” pull-down menu and choose “Select Library”

Click on the Browse icon and go to the C315 directory (e.g. W:) and select “NIST02.”

This will allow you to do your data searches. You will need to check with library personnel to see where printouts from PC 5 appear.

Quick Guide to Data Analysis on ChemStation: The left mouse button is used to zoom in on peaks. This feature works in either the chromatogram or mass spectrum windows. You draw a box around the area you want to zoom in on. The Y axis does NOT autoscale in the chromatogram window. Whatever is in the box is what fills the chromatogram window after the zoom. I suggest you start your zoom boxes from the bottom of the chromatogram. This way you can be sure that you do not cut off the tops of peaks. Double left-clicking will undo a previous zoom. If you want to redraw and entire chromatogram, click on the Draw Chromatogram icon (the icon looks like a little teal chromatogram with no arrows, letters, or other stuff in it). The GC method (containing temperature program, split ratio, mass spec conditions, etc.) can be found by selecting “List Header…” from the “File” menu. The instrument parameters are found in the “acqmeth.txt” window. The “Copy Window…” item under the “Tools” menu can be used for copying chromatograms and/or spectra into other programs. Window 1 is mass spectrum, Window 2 is the chromatogram. You can direct the computer to integrate the chromatogram by clicking on the integrate icon. It looks like a little blue chromatogram with a single, downward pointing arrow on it. Alternatively you can select “Integrate” from the Chromatogram menu. The results of the integration are viewed by selecting the “Integration Results…” icon from the Chromatogram menu. The right mouse button is used to select the portion of the chromatogram from which you wish to extract mass spectra. I suggest you draw the extraction boxes INSIDE chromatographic peaks to avoid extracting too much noise. Double right-clicking on a mass spectrum searches it against the selected library. You should only search SCAN mode data and use the NIST02 library. The “Tabulate” option under the “Spectrum” menu generates a table of masses and intensities from the displayed mass spectrum. This table can then be copied and pasted into Excel for further analysis. ChemStation does have a rather complete set of online help documents which can be accessed from the Help menu. Direct any further questions to Jonathan Karty, jkarty@indiana.edu.