“‘1‘ ‘11 1 ~ ‘1 111111111 :3 5111.111, , «11:11 1‘ 111,1 1" 1 “1331‘ “1“ 1111 L‘} ‘4‘5‘ “ .1“! amt. ' arr; 3...... “2.5553397 1. . $56.3“? 53:5: 3%“ z :‘2‘ 4:57:13 A y!” -.., . Jt‘r. . -,1'V ‘5 1h: 1 ‘El ,11 .1“ 1 11 1111‘11111‘1111131 111,111"i 1 .1} 1 ‘1' ““ 1".1, 1111111f 1,1 ' [1111“ 3,3? { 34131111 ‘ “1‘11: 55% 15%; -1 1 . '31"; ,5 23?. ‘ 3-5-35“: ”3:5“ 1 111 1",! .1 if 11% 141‘ 1' '1 1-21.. & ‘ *7-53’: {INN .zr“..;__ ' . ‘ '4‘ ~’1.- «a 5:1”. ..m.... x; 4‘ - 4‘ ‘41‘, (1 . ‘ 1 “%‘1‘1{1"‘1% ‘ ‘)‘:- :1“ }1}I‘1.1 “11‘11’1 *1 ‘1 j 111 Wu“ 11‘ 1'1”“, .. ‘ ,. J“. “Jars:- 1553‘. 3.. - ‘ ‘ l‘ 5.1.. .1; 11?,“fl 1‘1 1.11:1: fikf A ' - ’v-‘L $1551? 1” -35: Nan-4&5; 1- - —, gfii; :15: 11%.. figzj‘ ' - .{f a“ ' .-!!.:;:_,_-. 1. w 1% 1.1111111:- W1: 1&1 1111111. 1-1111? 1" ~( 5151:11311’13‘311"‘1‘1‘1‘1}‘.‘M1' .111“‘«1«' ”1,. 1 ‘1. 111116551:- 1 1M 51%:"1'1 ‘ i'lpénh‘h‘i 1. ‘té 3"" , ‘ 3%“ 1i}; ‘5} '11 ”.4“; _. .. :5 1%.-..» ~ . ‘- v a 41:} .1 - 1 .1 . . tr“ c r:—«. ‘35: ~ .1 5' _. ‘ 4’- "(A “4‘? c—v’" a. . - 33$ 1 , . . r ‘X. “ ~“¢ . l 4"?“ f5, .1 . ‘fig: ::£ a Q? .11. ‘11‘ ‘1‘; fi‘fi‘“ 1‘11‘14‘1“ ‘1‘“. 11"; ‘3 111‘ ,. fl -.1. “.4- r- f 1 ”1‘11 WE“ ““1 1. ‘11 1““ 1‘1» .3323" «Gui—3‘ ’L 3““ ‘4... _' 399;, 52‘ $11. a 0‘ ‘1 " 0 .11 54.? :.1 3111131531 ‘91“ 3" a 111,411“ 1111131 13".“5113‘5 - «3155“ ‘4" $91.11“ 1 “11‘ ‘tfifi‘fi‘ ‘gjvm‘ifi'éb‘ .. 3‘35 5E1 ‘n’L‘: 5%?“1” 1.” .- u - Trig-'IJr.‘ " c1514.?“ . ‘1 11331.1“V1‘L14L2; a”? ‘z 1211315“ ' (.9 461 :3. 13911}: 11111111 !' 025' ' 111 we" ‘71; . {‘51: ~ ’ :7. :14. .';_-‘: ‘4 11113311“: :4 a? '1‘} TH 9 5515 321 93 10590 0686 llllllllllllllllllllllllllllllllllllllllllllllllllll 7" - This is to certify that the thesis entitled SCREENING METHODS FOR ORGANOPHOSPHORUS PESTICIDES USING TRIPLE QUADRUPOLE MASS SPECTROMETRY presented by Mark R. Bauer has been accepted towards fulfillment of the requirements for M.S. degree in Chemistry Um Major protessor Date 4 /:7 Z (/{4 0-7639 MSU is an Affirmative Action/Equal Opportunity Institution SCREENING METHODS FOR ORGANOPHOSPHORUS PESTICIDES USING TRIPLE QUADRUPOLE MASS SPECTROMETRY By Mark Richard Bauer A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1983 ABSTRACT SCREENING METHODS FOR ORGANOPHOSPHORUS PESTICIDES USING TRIPLE QUADRUPOLE MASS SPECTROMETRY By Mark Richard Bauer The detection and identification of organophosphorus pesticides is an increasingly important problem. Detection methods normally re- quire multiple extractions and long chromatographic separation times. The methods described here utilize the low energy CAD fragments of organophosphorus compounds to quickly screen for pesticides, their metabolites and decomposition products. Methanol is used as the chemical ionization reagent gas in order to produce only (M+H)+ ions in the source. The (M+H)+ ions from each pesticide class follow characteristic decomposition pathways in the collision cell that may be monitored by neutral loss and parent ion scans. For a complete screening analysis, the triple quadrupole mass spectrometer automatically runs through a set of neutral loss and parent scans. The method can be used to detect many common pesti- cides and related compounds at the ng level in complex mixtures. The compound specific parent-daughter pairs exploited in this technique can also be used for targeted analysis using reaction monitoring. This limits the number of compounds that can be detected at one time, but it also lowers detection limits correspondingly. These methods illustrate how MS/MS can advantageously use the results of fragment- ation studies for the direct analysis of entire classes of compounds. ACKNOHLEDGEMENTS I would like to express my appreciation to Proffesor Christie G. Enke for his confidence in me during this work , as well as his advice and assistance in the preparation of this thesis. I would also like to thank Michigan State University and the Office of Naval Research for the financial support they provided. Special thanks are in order to the other members of Dr. Enke's research group, both past and present, for their help in developing the instrumentation and systems support that make projects like this one possible. Anne Giordianni and Adam Schubert were especially helpful and I thank them for sharing their time and knowledge with me. I would also like to thank Milton Webber and the rest of the regulars at the E. Beak and C. Dome Bar and Grill for providing a pleasant atmosphere and working environment in the laboratory. Finially I offer my thanks and appreciation to my wife, Susan, for her constant understanding, support and love. TABLE OF CONTENTS Chapter 1: Background and Introduction Pesticide Development Structure of Organophosphorus Pesticides Toxicity Detection Methods Chapter 2: Mass Spectral Behavior of Organophosphorus Pesticides Introduction Phosphorodithioates Phosphates Phosphorothiolates Phosphorothioates CAD Spectra of Organophosphorus Pesticides Phosphorodithioates Phosphates Phosphorothiolates Phosphorothioates Chapter 3: Comparison of Ionization Techniques Introduction Electron Impact Ionization Chemical Ionization Ionization for MS/MS Experimental Results and Discussion Chapter 4: Detection of Organophosphorus Pesticides Introduction Experimental I Results and Discussion I Experimental II Results and Discussion II Conclusion References Appendix ._.I _| Nat—A LIST OF TABLES Chapter 2 2.1 Pesticides Used 31 Chapter A 4.1 Scans Useful For Screening 61 iv LIST OF FIGURES Chapter 1 1.1 Persistance of Some Pesticides 1.2 Common Organophosphours Pesticide Structures 1.3 Flow Chart For Multi Compound Analysis Chapter 2 2.1 Initail Fragmentation Routs For a General Organophosphorus Pesticide 2.2 General Structures of Four Organophosphorus Pesticides 2.3 Mass Spectrum and Fragmentation Routes of Malathion 2.4 Fragmentation Routes For Dimethoxy and Diethoxy Phosphates Chapter 3 3.1 Compounds Used For Ionization Experiments 3.2 Mass Spectra of Malathion 3.3 Mass Spectra of Dursban 3.4 Mass Spectra of Parathion 3.5 Mass Spectra of Phosmet 3.6 Mass Spectra of Phorate 3.7 Ammonia Chemical Ionization Spectra of Malathion 3.8 Background Spectrum of Methanol Reagent Gas Chapter 4 4.1 158 Neutral Loss 4.2 186 Neutral Loss 4.3 125 Parent Scan 4.4 Spectra Obtained Using Scanning Techniques 4.5 Multiple Reaction Monitoring Results 4.6 Multiple Reaction Monitoring Results (dilute sample) WNW Chapter 1 Background and Introduction E l' 'l D J ! The use of chemicals to control insects and other pests has become commonplace in the world today. The fact is that chemicals have become necessary to preserve the way of life most people are now accustomed to. There are 4.7 billion people in the world today and the population is increasing at the fastest rate ever. At the same time, less land is devoted to agriculture each year. The only way that worldwide starvation is prevented is by increasing the pro- ductivity per acre through the use of chemical fertilizers and pesti- cides. If pesticides were not used, statistics show that only 37% of the potatoes grown would survive, 78% of the cabbage cr0p would be destroyed and only 10% of apples and 9% of the peaches grown would reach the market (1% In a single season a house fly can, through seven genera- tions, spawn 3.5 x 1012 individual flies. The woolly apple aphid can go through 20 generations in a season. Two aphids can mushroom into a phenomenal 7.6 x 1030 aphids (1). These insects along with others can completely wipe out a crop if left unchecked. They are so prolific that even with the use of pesticides, one third of all crops are lost to pests and disease. Insects, rodents, weeds, microorganisms and parasites represent a real threat of disease and destruction to plants, stored food and other dry goods, as well as threaten the health of livestock and humans. Fortunately science has developed chemical pesticides to control many of these problems. The use of pesticides has continued to increase yearly (1). The high economic efficiency achievable with the use of pesticides in agriculture and other branches of the economy has favored a rapid development of these compounds by the chemical industry. There has been continuous change andznnprovement.in the assortment of pesticides available. In the middle of the nineteenth century the first pesticides used were powders such as sulfur for controlling mildew in vines, bordeaux mixture (copper sulphate solu- tion plus lime) for controlling downy mildew in a variety of crops and paris green (a mixture of cupric oxide and arsenic oxides) against Colorado beetles. Further advances came at the beginning of this century with the use of nicotine solutions against aphids and tar distillate sprays to control aphids and lichens on dormant fruit trees. Ferrous sulfate and sodium chlorate were introduced as general herbicides and dinitrophenols and cresols were used to control weeds in cereal crops. These crude solutions and mixtures of multiple compounds have since become known as "first generation" pesticides (2). In the 1940's a second generation of pesticides was devel- oped. These new compounds included the organochlorines, organophos- phoruses and carbonates. For many years the pesticides most commonly employed were organochlorines such as DDT, BHC, aldrin and dieldrin because they were so effective against a variety of problems and did not have to be reapplied very often. Recently however, this category of compounds has come under criticism and the organophosphoruses and carbonates have become the main pest control chemicals. The chemicals in this second generation of pesticides have been developed to the point where they are optimized for specific applications. Different compounds are used for their particular'effectiveness. as insecti- cides, herbicides (weed killers), rodenticides, molluscocides (snail killers), acaricides (spider-mite killers), nematodacides (worm kill- ers) or fungicides. There is now a third generation of pesticides under develop- ment. These compounds are even more specific and are safer for the environment. New nontoxic chemicals such as pheromones and sex attractants are intended to disrupt the reproductive cycle of the insects rather than kill them outright (1,3,4,5). This new breed of compounds is especially useful in urban areas where toxic chemicals pose the greatest risk. The major reason that organophosphorus pesticides have become so popular is that, unlike the organochlorine compounds, they are less persistent in the environment (1%. DDT and dieldrin, the most comnuniorganochlorines used, were in use for over thirty five years will continue to be a part of our environment for years to come. Organochlorines are very persistent, having low volatility and water solubility. They are however, very soluble in fats and hence they collect in the fatty tissue of higher animals. The extent to which organochlorine residues have accumulated in soils and their side effects on the ecosystem have been well studied and documented. The result is that even though organochlorines are very effective in controlling insects, the shorter lived organophosphorus compounds are now used for most problem pests. Figure 1.1 shows several pesticides and their effective life times (6,7). The organophosphorus esters are chemically unstable, susceptible to both acidic and alkaline hydrol- ysis by a variety of mechanisms which render the compounds harmless (1,8). The compounds are also more toxic than the organochlorines before they decompose. ‘Therefore less of the chemical needs to be applied. This makes it cheaper for the farmer and better for his land. In the same manner that the organophosphorus pesticides can be made specific for insects or weeds and not humans, they can also be made to be highly toxic to humans. By changing the molecular struct- ure certain organophosphorus pesticides can also be used as chemical warfare agents. Before World War II, Dr. Gerhard Schrader made the first lethal organophosphorous pesticide that later became Toban, the first phosphorus chemical warfare agent" Much of the original re- search on the chemicals that became pesticides was done by the Germans during World War II in order to find a nicotine substitute (2). The toxins they discovered instead have found a place in peaceful non— mammalian uses. chflmanmmsnhmfiesticides The organophosphorus compounds under discussion all have a similar structure. A generalized structure is shown in the following figure. X R,O\;‘.‘>\ RZO/ Y—Z R1 and R2 are usually the same and can be either an ethyl or a methyl group. X and Y can be either sulfur or oxygen and Z is a fairly good Aiiniiov 1391801013 p33081103 10° -V\/—{’M/ DDT 50 -I .1 '1 J ‘ 1 o U ' t I I l 1' 1 I I I I T ' V f l I I 100 .. CARBOFURAN q DASANIT 1 Lil 8 L 4,3L . . 1 1 4 fl '1 0 c I I r I I r r’l—1 I I r I I f r r 1 100 PARATHION _ PHARATE .7 5° .1 q d 1 ‘ 1 0 I I I f I I I I I l I l I I 1 fl I U I O 10 20 30 40 50 O 10 20 40 50 Weeks After Treatment Figure 1.1 Persistance of some Pesticides organic leaving group such as a substituted or heterocyclic ring, an aliphatic chain or an amine. In 1963 Schrader came up with a formula dictating what functionilities are necessary to assure biological activity in an organophosphorus pesticide of the following general structure: (3) R/ P‘ACY‘ "It is likely that a biologically active phosphoric acid ester will be attained when the following prerequisites are satisfied: Either sulfur or oxygen must be directly bound to the pentavalent phosphorus, R1 and R2 may be alkoxy groups, alkyl groups or amines; while the "acyl" may be represented by the anions of organic or inorganic acids such as fluorine, cyanate and thiocyanate or one of the other acidic com— pounds."(8) Schrader's definition of "acyl" was rather liberal and differ- ent from that which we use today, but he was the first to understand that the chemical mechanism of insecticidal action may depend on the phosphorylation of biological targets (8% There are six main groups of organophosphorus pesticides made by the different possible combinations of sulfur and oxygen. The different classes are shown in Figure'L2. Each group can also con- tain compounds with R groups which are either both methoxy or ethoxy. There are so many different types of organophosphorus pesticides because they can be manufactured to accomplish a specific mission. By changing the R groups or the type of acyl moiety used, specific in~ sects, plants or other pests can be controlled. The compound's 10:01! 2) \/ 13:0 13=o \ / “azw I O I '0: I ‘1’ mm mm :0 \/ m phosphate phosphorothlolote phosphorothloate phosphorodithioate phosphonate phosphomidate Figure 1.2 Common Organophosphorus Pesticide Structures 7 solubility, and hence persistence, may be tailored to last an entire growing season or be made such that produce can be sprayed one day and eaten safely the next. Certain structures are better if incorporated into the soil as herbicides; others are better if sprayed, others still are best if used as a systemic pesticide. Various forms of the molecule can be a deadly toxin to one species but perfectly harmless to another. This way the applied pesticide will kill the pest that tries to take a bite but will be harmless to birds or humans that may be hungry (4L I . 1 Organophosphorus pesticides are thought to work by the inhibition of enzymes involved in the functioning of nerves (1,2,4L In human cases the enzyme affected is acetylcholinesterase (AChE) which under normal circumstances catalyzes the hydrolysis of acetyl- choline. During normal nerve activity when electrical impulses reach a nerve synapse, acetylcholine is released and diffuses across the synapse. Upon contacting the receptive site on the other side of the synaptic gap, a new impulse is triggered. acetylcholinesterase helps to hydrolyze the acetylcholine in order to free the receptive site and allow fresh impulses to be received. The hydrolysis reaction is shown in equation 1A. 0 _. + I fiz<3 X(CH3 )3NCH2CH20 — C-CH3 WE} u )7 (CH3)3 NCHgCHz—OH + CH3C—OH ACETYLCHOLINE CHOLINE Equation 1.1 If the acetylcholinesterase is inhibited, the acetylcholine builds up at the synapse and the nerve becomes locked in the "on" position causing, for example, muscles to remain contracted. The actual mech— anism for the hydrolysis of acetylcholine involves the formation of an initial enzyme substrate complex (equation1.2) followed by the re- lease of choline and the formation of esterified enzyme (equation 1.3). _ ll AChE + X(CH3)3NCHZCH20—C—CH3 "3— o AChE . - - (CHslsNCHchzo—g-CHJ COMPLEX Equation 1.2 9 _ . COMPLEX —>AChE-C-CH3 + X(CH3)3NCH2CH2-OH CHOLINE Equation 1.3 The esterified enzyme is then hydrolyzed back to free enzyme and acetic acid (equation 1.4% H 1123C) ll AChE-C-CHa —-> AChE+CH3 C-OH Equation 1.4 The organophosphorus compounds are effective inhibitors because they are phosphorylating agents due to the good leaving group Y. They are able to chemically combine with the enzyme forming a stable phosphory- lated compound (equation 1.5% This new enzyme is unable to react with acetylcholine because it‘s active site is blocked and it's hydrolysis is very slow (equation 1.6% \II R0\ /x AChE+R :-p Y- z :: AChE-——/P\/ R0 R0 Y—Z COMPLEX OIR COMPLEX -—> AChE—1"=x + HY—Z OR Equation 1.5 OR x I H20 RO\ // AChE— P: x ——> AChE + P I erv / \ 0R s|ow R0 OH Equation 1.6 The factors affecting the inhibiting power of the organophosphorous pesticides include size, shape and polarity. These factors affect the relative stability of the phosphorylated enzyme. The nature of the leaving group also affects the reaction rate of the initial complex formation step (2% The exact structure of AChE is not yet known, nor are the structures of the enzymes employed by the different insects. The nervous system of insects is thought to be somewhat similar to the human system but the differences allow some organophosphorus compounds to affect insects preferentially. The organophosphorus pesticide that is able to fit into the enzyme active site of the insect is unable to fit into a human's enzyme active site, thus making the chemical safer to humans(2). The different physiologies of beings in the animal kingdom 1O cause another type of specificity. In some animals certain organo- phosphorous compounds are metabolized or hydrolyzed to harmless phos- phorus acids before they reach the enzyme site (2,8L In many cases metabolism oxidizes the organophosphorus pesticides, creating a form which is more susceptible to hydrolysis than the reduced version. In other cases however, the oxidized form is even more deadly (1,2L This in viva conversion is capable of converting an innocent compound that can be applied widely into a substance that is toxic to only certain targeted species. The toxicity of organophosphorus pesticides is measured in terms of a factor called LD50 which is the average minimum dosage, in mg/kg of body weight, required to kill 50% of the population tested (2), LDSO values are determined using a variety of laboratory animals and are assumed to be similar for humans. The means of application also needs to be specified since injected doses, inhaled dosages and skin adsorption (the most common means) all have different LD50 values. The LD50 of malathion, a very common general purpose insecti- cide, is >4 g/kg dermally (9). Malathion, which was introduced in 1950, is a fast acting pesticide that is widely used on vegetables, fruits and cereals, both in the field and in storage. It is the compound used on humans to control lice and crabs and is also the toxin employed in animal flea collars. In 1982 it was the insecticide used in California against the Mediterranean fruit fly. It was mixed with molasses and yeast as bait and sprayed in residential areas by helicopter (4% It was used because it would take several hundred grams of malathion to affect an average size person and small amounts are quickly metabolized in man. In contrast, the LD5o of the chemical warfare agents known as V—agents is about 0.03 mg/kg. It is estimated that 0.09 grams of a V-agent, a drop about the size of a pinhead, when placed on a man's skin will be lethal. It is not an easy death either. The first signs of organophosphorus pesticide poisoning are constricted pupils, blurred vision and pain behind the eyes. Subsequently, tightness of the chest with difficulty in breathing develops. This is followed by drooling, sweating, nausea, involuntary defecation or urination and eventually convulsions, coma and death by asphyxiation; all in less than ten minutes (2% Most compounds fall in between the two extremes of totally safe and deadly. Researchers, manufacturers, distributers and farmers are urged to use appropriate safety measures when handling and applying these very helpful, but potentially dangerous chemicals. Caution must be used, not only to protect humans, but also to insure that no other creatures in the ecosystem are unintentionally harmed. DQLQQLiQn.M§LhQQ§ For each different application of organophosphorus pesti- cides a detection method has been developed. Pesticide presence is monitored as they are manufactured, as they are applied, after they are applied and after they have claimed their victims. Analytical methodologies are also developed for new compounds or new applications during their experimental stage as an aid to the researchers so that they can analyze the pesticide's performance. Pesticide detection methods have also been developed to monitor their presence in differ- ent parts of the environment. When pesticides are applied most of it ends up where desired, either in the soil or on a plant, depending on the method of application and intended use. However, through rain, wind, evaporation, diffusion and sometimes human carelessness, the pesticide is partitioned between the air, water, the soil and the rest of the ecosystem (10,11L In the beginning of pesticide development, scientists were mainly concerned with the effectiveness of the compounds; how well they worked, how long they lasted and which application method worked best. During this phase, crude detection methods were often used. To see whether a certain variety of insect or plant was affected by the compound, they were simply exposed to the toxin and monitored to observe any affects. In his study of the persistence of certain pesticides in the soil Harris used cricket nymphs placed in a con- tainer with some soil. The number of nymphs that died was then plot- ted versus the weeks after application to obtain the results (6% For humans however, this type of method cannot be used. So once LDSO levels were determined using laboratory animals, methods had to be developed to quantitatively detect the actual compounds in various species and in the environment. Finding the small residual amounts of pesticide present in food, plant and soil matrices is a classic analytical problem and classic analytical procedures were developed for their detection. The first steps in soil and food analysis generally involve a lot of wet chemistry (12,13). The sample is first homogenized and then the pesticides are extracted out of the solids with appropriate solvents. The extract can then be filtered, buffered and subjected to several more extractions, dryings and evaporations. Column chromato- graphy is often used as a last cleanup step before final detection. 13 Gas chromatography is the traditional detection method, but other detectors such as mass spectrometry and thin-layer chromatography are also used. A general flow chart of a typical multi compound analysis is shown in Figure 1&L Depending on the target compound and the sample matrix, different procedures must be followed because of the differences in extractability, elution patterns in different absorbent solvent systems and the selectivity and sensitivity of the various filtration, cleanup and detection steps. Screening methods, used to qualitatively find which compounds may be present are usually dif- ferent from quantitative analyses.in that they can often be accom- plished with less chemical work up. The detection of pesticides in water samples involves many of the same steps outlined above. However, the concentrations of the target compounds in water are often very small and much smaller than the amounts found in many foods. Therefore large volumes of water rnust be sampled in order to get enough compound to detect. Processing water samples is also complicated by the fact that most pesticides are insoluble in water but are easily adsorbed by suspended organic matter in the sample. This necessitates detection techniques that address both adsorbed and dissolved pesticides (14). If the analyst is inter- ested in the total pesticide contribution to lake or stream water, both states of the pesticide must be accounted for. If drinking water which is supposedly cleaned and filtered is under investigation, then only the dissolved pesticide need be considered. The water to be ana- lysed is forced through a chemical trap or a physical filter, such as ea macroreticular resin like XAD-2 or XAD-4 in order to preconcentrate the sample (15%. The sample is eluted or extracted from the con- centrator and then sent through the steps in the scheme shown in 14 [SAMPUNG \L Homogenize [Extraction ~ ¢ FHter Buffer / EXtrOCt < J q‘V~ “J Clean up CMumn D Chronmfiography lEvoporoHon ‘l lRinse T Vt SEPARAHON TLC GC HPLC DETECTION GC [Detectors MS Enzymes UV—Vh lNMR Polarography Figure 1.3 Flow Chart For Multi Compound Analysis 15 Figure 1.3. Air samples pose similar problems. Particulate pesticides and pesticide vapors can be present in the air at the ng/m3 level or lower. Air samples must also be preconcentrated by traps or filters. Pesticide vapors are usually chemically caught in resins, impingers or gas debubblers, while airborn particulate matter is trapped in filters and porous substances (16-19). ‘The collection techniques employ large fans or pumps to sample 300 m3 or more of air needed.for analysis. The pesticides are then removed from the sampler with a soxhlet ex- tractor, cleaned, concentrated and analyzed with a gas chromatograph. Many of these methods are still used today but newer analy- tical techniques have been added to improve the methods. Lower de- tection limits for pesticides in air have been reached by using a combination of polyurethane foam and'Tenex resin preconcentrators (20). Tenex is much better at trapping organic vapors and hence the formerly "hard-to-capture" high volatility pesticides are effectively collected. Organic interferences can be removed by a chromatographic cleanup and fractionation step which also aids in lowering detection limits below 1 ng/m3 . Small portable air samplers have been de— veloped that do not need large air flow rates and are ideally suited to detect vapors in the work place (21% High Performance Liquid Chromatography (HPLC) has become a popular tool for water analysis (22—24%. Samples as small as 1 ml can be analyzed on chemically bonded stationary phases without any precon- centration, and using solvent programming, part per billion detection limits are realized. Because of its ability to concentrate and separ- ate, reversed phase liquid chromatography has also been used to detect organophosphorous pesticides in plasma and other body fluids. Recent 16 developments in HPLC detecters such as the LC/MS interface have given HPLC excellent sensitivity and selectivity for pesticide analysis (24). Thin layer chromatography (TLC) can also be used without anCh cleanup for a variety of messy samples (13). JIt is often used as a quick screening method to find which longer quantitative methods should be run on a sample. A new very selective method for organo- phosphorus pesticides has been developed which utilizes cholinesterase inhibition during the development of the TLC plates (24). Gas chromatography is still the most widely used detection device for organophosphorus pesticides (13,23-26). Most of the methods used for detection of organophosphorus pesticides on food (25,26) and in blood or urine (27), as well as the techniques men- tioned above for air and water analysis involve gas chromatography. All the compounds can be detected with one or another form of gas chromatography. All the latest advances in CC technology have been incorporated into the standard detection schemes allowing better separations and sensitivity. GC/MS using selected ion monitoring has been shown to be an extremely sensitive technique with detection limits less than 1 ppb (28). There are several inherent limitations in all of these detection methods, the most obvious of which is time. The time for one analysis can reach into days. With 24 hour collection times, 16 hour extractions, several hours of cleaning, evaporation and finally gas chromatography, rapid real time results are not possible. Fur— thermore, all the steps involved in the analysis cause errors due to incomplete extraction, adsorption on container walls and other solids and loss of sample during transfers. This thesis project is part of a 17 larger research effort to develop a method to detect organophosphorus pesticides in the air on a completly automated real time basis. None of the methods outlined above would be suitable to continuous monitor- ing or be easily automated. It will be shown in later chapters how triple quadrupole mass spectrometry can be used to detect organophos- phorous pesticides . Qaapter 2 Mass Spectral Behavior of Organophosphorus Pesticides W The mass spectrometer has become an important tool in the detection of organophosphorus pesticides. Unequivocal identification of pesticides and their metabolites in most cases requires the use of two analytical techniques (29). The retention time of a gas chroma- tographic peak does not provide enough information for absolute ident- ification (30). Comparison of the retention times from two columns with different characteristics to pesticide standards or combined data from chromatography and an optical technique such as IR or UV-VIS can be used if the pesticide has been identified earlier and reference data are available for comparison. These techniques are not as use- ful, however, when looking for pesticide transformations during meta- bolism and decomposition studies because the altered compounds may not have reference data available for comparison. A mass spectrum does contain enough definitive information to confirm and identify the presence of organophosphorus pesticide residues. Mass spectrometry fragmentation patterns resulting from bond fission and rearrangements are diagnostic and characteristic of the original compound. Therefore, by correlating the fragmentation patterns, pathways and mechanisms with the structure of known pure pesticidal compounds, the structure of unknown compounds can be de- 19 duced from their mass spectra. The theories and principles of the interpretation process in mass spectrometry has been discussed in sev- eral text books (31-33). In order to use mass spectrometry to detect the many forms of organophosphorus pesticides, studies must first be made of the spectra of pesticides with known structures. Several researchers have studied the fragmentation of organophosphorus pesticides during their development of GC/MS or straight MS analysis methods (34-36). A general fragmentation scheme is shown in Figure 2.1 which compiles the possible ion-decomposition pathways deduced by the pioneers in organo- phosphorus mass spectrometry and given in several reviews on the subject (37.38% There are four different types of organophosphorus compounds, made by different combinations of oxygen and sulphur in the X and Y positions. The structure of the phosphate, phosphorthioate, phos- phorothiolate and phosphorodithioate are shown in Figure 2.2 . The relative intensities of the ions formed by each of the different decomposition pathways depicted in Figure 2.1 are dependent on the type of organophosphorus pesticide present, but are in general governed by five factors: (31,37) 1. Stability of the ion produced 2. Stability of the neutral lost 3. Bond lability 4. Steric factors 5. Inability of ions to refragment The mass spectra of the four different types of organo- phosphorus pesticides do not have many features in common. Because the Z moiety is a good leaving group and often contains easily ionized 2O Rg3P=X R0. I“ R RO,P=XH 4 ZY" Pl m A ZYH‘ 2 __. X +. —l+ 214' 1N:3 ‘__ F21:)\~g; TT1C)\\ /’ )< ‘-__-—__T" F)\. “W Ro’ \Y—z P3 R0” Y __1* l R0\ /XH ‘+ P \ RO\ ,> | / 2 2 2 3 CH30 CI{28942 P4 type ions can be formed by a McLafferty type hydrogen rearrangement with charge retention. Examples of this are the ions at m/z 158 or 186 for corresponding dimethoxy and diethoxy ions, in the spectra of malathion and disulfoton (36% 23 + + CH 3,o\ P/SH CZH50\P/SH s / \ CH30/ 5 C2 H50 5 Azinphos, which does not have the gamma hydrogens needed for McLaffer- ty rearrangements makes P3 ions at m/z 157 (36). Although a double hydrogen rearrangement is theoretically possible for some compounds, such asxnalathion, ions from this mechanism are not very abundant. Phosphorus containing ions of the P1 type are also found in most dithioates at m/z 125 for dimethoxy and 153 for diethoxy compounds, but at much lower abundance than the N type ions (34). The N ions can fragment further producing other ions indicative of the structure of the Z group. The mass spectrum of malathion and its fragmentation scheme are shown together in Figure 2.3. The only unusual peak in the spectra of the dithioates is the base peak in the dimethoate spectrum at m/z 87. High resolution Ineasurements gave a formula consistent with the structure shown below, but the mechanism of this fragmentation has not been explained (34). “Ii.” —]+ CHBO g > swaying—CH, —~ S‘CHz“ 0- N-CHs CH30 H m/z 229 m/z 87 Phosphates The mass spectra of organophosphates can also be dominated by ion produced by the cleavage of the Y-Z bond, but unlike the 24 100 - 50'- It. 0 17 I U ' r r I I If r r I I I‘ I l I I I l 60 100 140 180 220 250 300 340 S 0 H P-S-CH - C-O-C2H5 CHSO/ CH C o C H 2' " " 2 5 / 3 \ o 33 II CH30\ ,SI-I —9 fiHC-O-Csz p CH30/ \s CH30\.S. l 9 CHC-O-Czl-ls L5_§ P-S-CH-C\ 3 m, 0130’ I 00sz o CHEfi/ II CH C \ CH-C 3 \st 0 II \002H5 CH30/ 2.5. CH-C/ as \ 3 I31 8 CH30 CH3 CH-C \P \stw II :0 CH30/ 15 CPI-'0' 2.3. l o 22 H0 \P ——> P-OCH; CH-C: CH3O/ 6—2- 3" 0 1—9- A Figure 2.3 Mass spectrum and fragmentation routes of malathion 25 dithioates, the phosphorus fragment retains the charge. Double hydrogen rearrangements are often prevalent, producing P5 ions. This mechanism was proposed by Domico for the formation of the base peak at m/z 127 for phosdin and phosphamidon (311). +l—— —_1 + CH30\C") 9 CH3O\9 0113 C") /C2H5 /P-O-—C=CHC——OCH3 /P—O—C= ('3-C-N\ CHSO CH3 CHSO / Cl 0sz \ + CH30\P/OH / \ When hydrogens are not available for rearrangement, P1 ions are gen- erally produced. The base peak in the spectrum of the oxygen analog of diazinon, at m/z 137, is due to a P1 type ion (37). +~ — + 1 C H 0 02H50\|FO|, Oh 2 5 \ CZHSO CH(CH3)2 Cszo m/z I37 Single hydrogen rearrangements are not observed in the initial fragments of the Y-Z bond but the P1 and P5 ions fragment further with successive single hydrogen rearrangements, losing ethylene (29). Fig- ure 2.4 shows the general scheme for dimethoxy and diethoxy phosphates (36). The ions in Figure 2.4 are responsible for the majority of the peaks in the mass spectra of the phosphates, but there are generally a 26 CHao\9 cuso/ Fo/m \m CH30 \ /0 CH30\ —> P=O CH30/ \OH CH30/ l2? ‘ |09 CH30 :P CH30 93 C2H50\\9 P-O-Z Cszo// Ps/ \PI CszO\ P/OH CHscH20\ / P\ —> /p:() CszO 0H5 CH3 CH20 1:5 I I37 CszO H—> \on CzH50\ Ho/ P:OH l l2? HO\ OH HO P/ ——-> \on Ho/ \0H 99 8| Figure 2.“ Fragmentation Routes for Dimethoxy and Diethoxy Phosphates few small peaks due to the Z moiety. N2 type ions are typical and are useful in conjunction with the molecular ions and P1 ions in ident- ifying the formula of phosphates (37). lflxuuflmuihiclates This group of compounds has the fewest general character- istics of any group. Each compound seems to undergo its own reac— tions. The N ions from the Z moiety dominate most of the spectra, but they are not formed by consistent mechanisms. Many of the ions formed have unexplained reactions. Exact mass measurements indicate the elemental composition of ions that use single or double hydrogen rearrangements, but where these hydrogens come from and how they migrate is not always obvious. An example of this is in the spectrum of the oxygen analog of dimethoate, in which the acidic amine hydro- gen is postulated to be responsible for the m/z 156 ion (311,38). T .1. /P-S-CH2- C—N —_. /p\ CH30 CH3 CH3O s-CH2 111/2 2|3 m/z |56 Fragmentation can occur on either side of the sulphur atom, producing P1, P3, N1 and N3 ions as well as the associated hydrogen rearrange- ment ions. The dimethoxy compounds generally produce at least some ion current due to the ions at m/z 109 and 79 (34,36). The 109 ion could be formed as a P1 ion or from the loss of H23 from a p5 ion (36,38). The ion at m/z 79 is from the loss of CH20 from the m/z 109 ion with hydrogen rearrangement (38). This reaction is preferred over 28 loss 0f CH30- because CH20 is a more stable leaving group (37). 4. \§ /. CH30\ /OH—] CH3O P-S-Z P CH30/ CH3O/ \SH m/z I43 + + CH3O\ —[ _l /P—O ——>CH30-P-OH CH30 m/z 79 m/z l09 We. The phosphorothioates are characterized by a rearrange- ment of the Z group from the oxygen to the sulphur (29,36,38-40). The migration is either initiated through thermal isomerization prior to ionization or by the electron impact process itself. Cleavage of the phosphorus-sulphur bond is preferred over Cleavage of the phos- phorus-oxygen bond because sulphur contributes less to the pi bonding (8). The ions at m/z 109 and 79 arise after the rearrangement(38). j‘“ 4' R0 9 R0 9 CH30 *1 >I’3—o—z——~ >P-S—Z-—-’r >P=O ——»CH,,o—P—0H R0 R0 CH30 m/z 79 m/z 109 Other compounds do not rearrange as above but fragment with double hydrogen rearrangements making P5 ions (36). Many of the diethoxy compounds show a loss of one of the R groups as CZHu (3“.37). This rearrangement has been named after Quale who first observed the loss 29 of the stable ethylene neutral from trialkyl phosphates (in). +- + fi ——) I10 —1 —z P-X—Z -———+ The ethyl group can also migrate to the Z moiety during fragmentation CH3CH20 \ I \\ // —Y CchHZO CH3CH20 as is seen in the m/z 179 ion in diazinon (311). +- + CZI-go CH(CH)—l CH(CH ) l N— 32 32 S 4\ 91% ,3- C2H50 CH3 CH3 m/z l79 EMQWW In the same way that correlations between fragment pat- terns and structures of compounds must be made before the spectra can be of use in analytical techniques involving a mass spectrometer, correlations between daughter spectra and parent ions must be made before they can be used in MS/ MS techniques. This section reports the collisionally activated dissociation (CAD) spectra of the (M+H)+ ion of 37 organophosphorus pesticides obtained on a triple quadrupole mass spectrometer. Table 2.1 lists the 37 pesticides by classes, with their structures and molecular weights, as well as the appendix page where the spectrum for each compound is presented. 30 Table 2.1 Pesticides Used W azinphos dimethoate malathion phosmet thiometon W azinphos ethyl dialifor disulfoton dioxathion ethion phorate M11 317 229 330 317 246 345 393.5 274 456 384.5 260 19:91:12 ll ”FWD NQN 0 ~sz C NH CH3 0 ‘0.” C 009‘s CH2 goczl-I5 O 2 N: : O -CH2CHZSCHZCH3 -CH2 CH2 SCHZCHS § ”1 (C2H50)2P sAo s —CH,s P(OC2H,)2 31 Amendixfiase 1O 21 29 37 13 12 17 W dicrotophos heptenophos naled phosphamidon tetrachlorvinphos D HO chlorfenvinphos oxygen analog of dursban W oxygen analog of dimethoate oxygen analog of malathion oxydemeton oxygen analog of phosmet Table 2.1 (cont'd) 237 251 381 299 366 359.5 333 213 314 246 301 32 o -C- CHC-N(CH3)2 CH3 73:) Cl —C'H - C'(C|)a Br Br 0 -q-CCI— 6 -N(CZH5) CII3 2 ClCH Cl -"Q. CI CI ‘1? Q. Cl CH Cl QC. 0 —CH2CNHCH3 —CH 2 o CZH5 CH2? 0 C915 0 0 —CH2 CH2 3 CH2 CH3 -CH2—N \ O 20 23 31 36 16 11 22 2M 30 W demeton-S oxygen analog of phorate W methyl dursban etrimphos fentrothion parathion-methyl pirimiphos—methyl temphos W aspon demeton-O diazinon dursban 258 244 323 292 277 263 305 466 322 258 304 351 Table 2.1 (cont'd) 33 -CH2 CHZS CH2 CH3 —CH,s CH2 CH3 HZCH3 N=4§ Kl” C HZCH3 CH3 CH3 F»: \,.—1 N(CZH5)2 3 03-00 (5(0CH3), 2°? — P(OC2H5)2 — C HZCst CHZCH3 N CH(CH3)2 fiN \\ I / CH3 C «5 >.. w Cl 28 15 18 19 26 33 35 14 parathion pirimiphos quinalphos Table 2.1 (cont'd) 291 25 Om. 333 / \N 32 298 N 3n END 34 The spectra were not all taken at one time, and therefore were run under slightly different instrument conditions. Liquid sam- ples were placed in a capillary tube. Solids were either put inza capillary tube also or dissolved in ether and then evaporated onto a quartz rod. All the samples were then introduced into the mass spec- trometer by means of a solids probe. Methane, isobutane or methanol was used as the chemical ionization reagent gas. The argon collision gas pressure was adjusted until the molecular ion was no longer the base peak but was still present in the spectrum if possible. All spectra were taken with.20 volts of collision energy. Some of the spectral data were recorded with as oscillographic recorder and some were taken with an automated data system. Instrumental conditions can affect the the ion intensities but do not change which ions are present in the spectrum. These are representative spectra of each compound but cannot be used for quanti- tative comparisons because identical instrument "tunes" could not be used during the acquisition of data. In general the CAD spectra have fewer peaks than the corresponding EI or CI spectra. For many of the compounds, the peaks that are present in the CAD spectrum are the same»as those found in the BI or CI spectrum. Other compounds however, have very different spectra. Ions that are formed by the CAD process may not be present in the normal spectrum and many of the ions from the BI spectrum, including abundant ones, are not produced by collision. This is due to the different energies involved in the collision and ionization pro- cesses . 35 El l 1'“ . I Every one of the dimethoxy dithioate spectra contained a P1 ion at m/z 125 except phosmet. The only daughter ion, and the base peak, in the phosmet spectra.is due to an N3 ion. The base peak of the azinphos spectrum is also an N3 ion at m/z 160. An N3 ion is also responsible for the base peak at m/z 89 for thiometon. This ion probably has a protonated tetrahydrothiophene structure. 7’“ 4' CHaegis-mu CH—s—CH CH -—-—- C'Hz 0'42 /P 2 2 2 3 CH2\S/CHZ m/z 89 The dimethoate spectrum has a base peak due to loss of 31 mass units which is probably loss of a methoxy group. The ion at m/z 89 is probably the protonated version of the rearrangement proposed by Domico as discussed above (34% The spectrum of malathion shows a base peak at m/z 127. 'This ion as well as the ion at m/z 99 are refragmentation products of an N4 ion following the same fragmentation path outlined above. The mass spectrum of disulfoton, the diethoxy complement of thiometon, also contains only one peak. The N3 ion is the same as the one in thiometon. The mass spectrum of phorate also has only one peak and it too is an N3 ion. The base peak for azinphos ethyl at m/z 160 is the same, as expected, as the base peak in the azinphos spectra. This ion can then lose C=O to yield the ion at m/z 132 (34). 36 _]H+ 0251-103, + \ I P--S CH -N :0 CH N ASWD CZHSO N. / m/zI NC T W2 |32 The spectrum for azinphos ethyl as well as those for ethion and dioxa- thion show P1 ions and daughters of P1 ions. The peaks in all three compounds at m/z 153, 125 and 97 have the structures + —,+— + C2H50\ —] HO\ HO\ ‘1 -P=S——> P=S ——+ P=S CZHp/ CZHSO/ HO/ m/z 153 m/z I25 m/z 97 The base peak of the ethion spectrum is caused by a cleavage between the carbon and either sulphur and, because of the symmetry in the compound, it is both a P1 type Z ion and a phosphorus ion. The ion at m/z 142 is the result of a rearrangement and loss of an ethoxy group. +|~-|+ + C2H50\s s/oc2 H1 :1 _. S /P- -—s CH— —s— CH5— 3— I'>\ C2H50\P< CZH50 OCZHs m/z I42 The peak at m/z 271 in the spectrum of dioxathion is also a Z ion and a phosphorus ion caused by the loss of one of the dithioate esters. 37 +H + C2H50\o 7‘ P—S o + / CZ H5 0 1 C2 H50\IS Cszo —S O \ _ cszo/ 7 S O m/z 271 The only peaks that were identified in the CAD spectrum of dialifor are at m/z 186 and 208. They are the complementary P3 and N3 ions. Phosphates Few of the phosphate spectra contained ions due to the Z group. Because the Z moieties of most of the phosphates are halogen- ated aromatic groups, N type ions predominate. The exceptions are posphamidon and dicrotophos, both of which have P3 ions forming the base peak. Dicrotophos also has a peak at m/z 72 which is probably formed by alpha cleavage at the carbonyl group. +H+ + CH3O o o —] o _] \ I CH CH g-O—C-CH-C—N< 3 ——» C—N< 3 CHSO/ CH3 CH3 CH3 m/z 72 All the dimethoxy phosphates show an ion at m/z 127 which is the P5 ion. This peak is the only one present in the spectra of both naled and tetrachlor. 127 is the base peak in the spectrum of heptenophos which also shows a large peak at m/z 125 due to P3 ions. 38 The spectrum of chlorfenvinphos also contains a peak at m/z 127 but it is the product of a Quale rearrangement of the P5 ion at m/z 155. There is also a small peak at m/z 205 which is due to some N3 ions. The oxygen analog of dursban does not fragment by losing the Z group. It undergoes two Quale rearrangements making ions at m/z 3 06 and 278. W The spectra of the phosphorothiolates were very similar to their dithioate counterparts. N3 ions which are postulated to have the same structures that are described in the dithioate section, were the base peaks in demeton-S (oxidized disulfoton), the oxygen analog of phorate and the oxygen analog of phosmet. The oxygen analog of demeton forms an ion by alpha cleavage at the sulfoxide to yield m/z 169. Its other peak at m/z 105 is an N3 ion. The spectrum of the oxygen analog of malathion also has the same peaks as the malathion spectrum. The oxygen analog of dimethoate, like dimethoate does not follow the same pattern as the other compounds in this class. Nor does it behave in the same manner as dimethoate. The m/z 88 ion, which is the base peak in dimethoate, is barely discernable. The spectrum of the oxygen analog has two P1 ions; one at m/z 125 is due to a re— arrangement of the Z group on to the oxygen and then fragmentation, the other is due to the unrearranged species. The ions at m/z 155 are probable due to alpha cleavage at the carbonyl and m/z 180 ions are the molecular ion minus one methoxy group, as shown below. 39 m/z 155 W Many of the CAD spectra of the (M+I-I)+ ion of the thioates contain peaks due to Quale rearrangements. These are the only peaks observed in the spectrum of parathion. The spectra of demeton-0 has only one peak and it is from an N3 ion that has the same structure as demeton-S. Rearrangements of the Z moiety from the oxygen to the sulphur give rise to the two prominent peaks in the spectra of pirimi- phos. The m/z ion at 197 is for (P-Z)+, while the ion at m/z 181 is for (0-Z)+. The spectra of diazinon, Quinalphos and Dursban all con- tain P1 or P5 ions and their fragments. wa/ILCT “(K/I347] -———’HO\P=S—] / C2H50/ 021-150/ H0 "V2 |53 m/z |25 m/z 97 The base peak of dursban is created from N3 ions. However the spec- trum of methyl dursban does not have any N3 ions. Instead it has a base peak at m/z 125, a P1 ion that is characteristic of the dimethoxy compounds. P1 ions are also the base peak in etrimphos, temphos, methyl parathion and in its oxygen form in fentrothion. Fentrothion and methyl parathion show both forms of the Z group rearrangement. The base peak in the spectrum of temphos is due to P1 ions 40 ‘which is not surprising since there are two thioate groups in the compound. The peak at m/z 244 is due to cleavage on either side of the central sulphur, and 341 is really an N3 ion. The spectrum of methyl pirimiphos also contains the P1 ions at m/z 109 and 125 as well as a base peak due to N5 ions. 41 Chapter 3 Comparison of Ionization Techniques W The choice of ionization method can be crucial to the success of mixture analysis with a mass spectrometer. The sensitivity and selectivity of the analysis can be greatly Changed by the ioniza- tion technique employed. Many different types of ionization have been developed, each having advantages for the many different types of samples assayed with a mass spectrometer. Some ionization techniques ionize every compound in the sample, others ionize only the target molecules and leave the rest of the sample matrix alone. The tradi- tional method, electron impact ionization (ET), is very energetic, causes many fragmentations and yields a mass spectrum with many peaks. The fraction of unfragmented molecular ions is often so low that the molecular weight of the compound cannot be determined. Chemical ionization (CI), however, can be used to increase the relative amount of molecular ion present in a spectrum. (M+H)+ protonated molecular ions are formed during an acid base reaction in the gas phase between an acid reagent ions and a basic reactant molecule. BH" + M = B + MH" The exothermicity of the reaction depends on the relative proton affinities of the acid and the base. Any excess energy in the reac- tion is available as internal energy to the (M+H)"’ ion and promotes fragmentation. For maximum ionization efficiency one needs as exo 42 thermic a protonation as possible. But, to maximize the probability of (M+H)‘+ ion formation, which provides molecular weight information, one needs the least exothermic reaction possible. On the other hand, the fragment ions, which provide information useful for structure elucidation, need excess energy for their formation. If the sample is a mixture, then fragmentation should be avoided because they may obscure (M-I-H)‘+ ions of lower molecular wieght. Therefore, depending on the type of sample under study, different chemical ionization reagents can be tailored to give the right amount of excess energy to yield an adequate amount of (M+H)‘+ ions and fragment ions. .EhanulelmDflQLJExflUEEJQn To date several different methods of ionization have been used to ionize organophosphorous pesticide samples. In 1966 Domico (34) used EI to study fragmentation and rearrangements in 23 organo- phosphorous pesticides because it was the predominant method avail- able at the time. Domico was chiefly interested in the mechanisms and rearrangements involved in the fragmentation of organophosphorous esters. Basing his findings on work done by Quale (41), Domico pub- lished the first spectra of organophosphorous pesticides and explained many of their prominent peaks. The spectra were characteristic of E1 spectra with little molecular ion and many lower mass fragment peaks, and have become the reference spectra for organophosphorous pesti- cides. While Domico's work is very important in understanding the behavior of organophosphorous ions in a mass spectrometer, his study is not useful as an analytical technique. HuLStan (36) however, was 43 able to use EI mass spectroscopy in conjunction with gas chromotogra- phy to detect organophosphorous pesticide residues on food. Stan studied the four groups of pesticides and found five ions that were characteristic of the different classes. The ratio of the ions at m/z 93, 97, 109, 121 and 125 were found to be indicative of the type of organophosphorous pesticide present in a GC peak. These ions, along with the retention time, enable positive identification of the pesti- cides in food residues. C] . 1 I . l . Several people have also tried chemical ionization to analyze for organophosphorous pesticides. Holmstead and Cassida pub- lished the first CI spectra of organophosphorous pesticides in 1974 (42). They used methane as a reagent gas but found found the resul- ting spectra contained less (M+H)+ ion than the M" in the El spectra. The spectra contained some new ions, but lacked many of the El ions, including some that formed base peaks. They suggested that a softer ionization gas, such as ammonia, be used to increase the amount of (M+H)+ formed. Stan also investigated the use of chemical ionization for organophosphorous pesticides (43). He compared the mass spectra obtained using methane, isobutane and methanol as reagent gases. Stan found that chemical ionization gave better results than electron impact. Unlike Holmstead and Cassida, Stan had more (M+H)“' ion with methane than with EI, but he got even more with isobutane and still more with methanol. He suggested the use of isobutane because it gave, in almost every case, (M+H)+ as the base peak, as well as some 44 useful fragment ions. Busch etal. (44) also compared five ionization methods on several organophosphorous compounds. They compared EI, methane CI and three negative ion methods and found that none of the five techniques produced a satisfactory abundance of molecular or (M+H)"' ions. Negative chemical ionization of organophosphorus pest- icides was also investigated in several studies. Daughtery and Wander (45) used dichloromethane as a reagent gas and formed characteristic (M+Cl)"‘ ions. However, these ions were small and the spectra were dominated by (M-Z)"’ ions. Rankin (46) studied the negative ion spec- tra of pesticides including some organophosphorus compounds and he too found little or no molecular ions. Stan and Kellner (47) used methane electron capture in the analysis of the negative ion spectra of 52 organophosphorus pesti- cides. The small amount of molecular ion formed was dependent on the structure of the compound and its ability to stabilize the charge. They found that fragmentation always led to a few intense ions charac- teristic of the structural group of organophosphorus pesticide. Selected ion monitoring of these ions in GC/MS were used to determine pesticides in food and biological and environmental materials. W In an MS/MS experiment the ionization criteria depend on the purpose of the investigation. MS/MS is used primarily for either structure elucidation or for the analysis of mixtures. Structure elucidation experiments need fragments formed in the source to produce parent ions representing as many parts of the compound as possible. 45 Structure experiments are conducted with pure compounds, therefore there is no possibility of assigning a fragment to the wrong compound, or masking a lower weight molecular ion. The most important consideration in TQMS mixtureIanalysis is the amount of molecular ions formed. By keeping as much of the ion intensity concentrated in the molecular ion as possible, the sensi- tivity of the technique is optimized because the intensity of the daughter ion spectrum is maximized. Creating only molecular ions in the source also eliminates the possible confusion between fragments formed in the source and molecular ions of the same mass. The ability of'the first quadrupole to select a single mass makes it easier to analyze mixtures by automatically subtracting the background peaks as well as all the peaks from other components in the mixture. There- fore, the ionization technique does not have to ionize only the species of interest but can ionize all compounds in the sample without any ill effects. W]. In this study six different ionization techniques were tested to find the technique that was best able to ionize organophos- phorous pesticides for use in triple quadrupole mass spectrometry. The methods of ionization used were electron impact and positive chemical ionization using methane, isobutane, methanol, water and ammonia as reagent gases. Five representative organophosphorous com- pounds were chosen to test the different techniques. The set of compounds shown in Figure 3.1 contain dithioates and thioates, methyl and ethyl esters with aromatic and aliphatic:Z groups containing a 46 S s f? CH3'»‘°\|| II - P-S-CH-C-O-CHa-CHs CH3 °\” /° CHs—O/ ' /P-S-Cl-l2-N\ CH-fi—O-CHz-CHs CH3'0 3 O MALATHION MW 330 PHOSMET Mw 3|? CH=.I—CH2-o\,5| P-S-CHz-S-CHZ-CHg, CH3-CH2-O/ PHORATE Mw 260 S 3 CL CHs'CH2'0\|| CHs‘CHz-O\|| —— P.0‘.’N02 P‘O \ / CL CHs-Cl-Iz-O/ CH3—CH2-O/ CL PARATHION MW 29l DURSBAN MW 351 Figure 3.1 Compounds used for Ionization Experiments 47 variety of functional groups and hetero-atoms. The pesticides were obtained from Chem Service, Inc. and were used without further purif- ication. Liquid and solid samples were placed in melting point capillaries which were then inserted into the mass spectrometer by means of the solids probe. Methane, isobutane and anhydrous ammonia specified as instrument grade or better were obtained from Matheson and introduced into the source using standard procedures. Absolute methyl alcohol was purchased from Mallinckrodt and the Milli 0 water from Milli Pore. For use, approximately 3 ml of the liquids were injected with a syringe into an evacuated 250 ml expansion flask. Their vapors were then leaked into the source through a needle valve. Spectra were taken on an Extra Nuclear ELQ-400-3 triple quadrupole mass spectrometer which has an automated data system. The spectra were obtained by scanning the first quadrupole with the other two quadru— poles in the RF-only mode. There was no collision gas present in the second quadrupole. EI spectra were taken with 70 electron volt elec- trons, CI spectra were taken with 300 electron volt electrons. The CI gas pressures were varied for optimum ion intensities and were measured with an ion gauge attached to the ion source vacuum chamber. W The electron impict spectra were similar to those published by DomicoI(34) and also by Busch etall44) and contained many typical electron impact ionization features. There was little or no molecular ion present but many fragments were formed, especially at lower masses. The spectra can have many high intensity peaks, as in mala- thion (Figure 3.2), dursban (Figure 3.3) and parathion (Figure 3.4), 48 100 -- El so- .1 o L L I I I I I I I I I I r I ‘°°" Methane so- 0 ‘leLAAALH— .LL [L _L I I I I I I I I I I I I I I 10° "1 lsobutane 50- 1 l l ° IiirdI‘IIIIIlluIWTII “3°" Ammonia 50d 1 01 1°°1 Methanol 4 501 1+ OTIIIITIITITIFZLIII 1°01 Water 50-I 0 J_- l J in IIIfIIIIIIIIITII 60 80 100 120 14-0 160 180 200 220 240 260 280 300 320 340 360 380 Figure 3.2 Mass Spectra of Malathion 49 100 50 100 50 100 50 100 50 100 50 100 50 El i - Methane 11 -l .. hi .2 I I I I I I I I I ‘ lsobutane d I I I I I I “I I I I I I “ Ammonia -I q .A A l A; ILL A. I I I I I I "I I I I I I "" Methanol L I I I I I I I I I I I I I I " Water q -i I I I I I I I I I I T r so so 100120140160180 200 220 240 260 230 300 320 340 360 380 400 Figure 3.3 Mass Spectra of Dursban 50 100 50 100 50 100 50 100 50 100 50 100 50 '3 El 1 'I I I I ‘ Methane ‘ Methanol Ammonia WIT‘I *I‘ Wiring—r. k A l IIIIIIIIIIILIII so 80100120140160180 200 220 240 260 280 300 320 340 Figure 3.4 Mass Spectra of Parathion 51 or can be dominated by one peak from an especially stable ion like the m/z 160 ion peak in the phosmet (Figure 3.5) spectra or the m/z 75 ion from phorate (Figure 3.6). The spectra contain both fragments from the 2 group, such as m/z 173, 127 and 99 in the malathion spectrum, and fragments from the phosphorous esters ie. m/z 158, 125 and 93 in the malathion spectrum. Because there are so many peaks and many of them are involved in multiple fragmentations and rearrangements, there are not many useful generalizations that can be drawn from these spectra. The methane and isobutane spectra were similar in that there was more molecular ion present and fewer fragments than the E1 spectra. ‘The fragment ions still carried more of the ion current than ‘was desired. Both gases gave (M-I-H)+ ions with enough abundance to get good daughter spectra but were also energetic enough to cause frag- mentation. lsobutane with a proton affinity of 196.0 kcal/mole is a much "softer" ionizing agent than methane which has a proton affinity of 130.0 kcal/mole. The isobutane spectra therefore, as expected, had a greater percent (M+H)+ ion and fewer fragment ions. The base peaks in the isobutane spectra were always the~(M+I-l)+ ion but there were also a few other peaks present” The base peaks in the methane spectra were sometimes the (M+H)"’ ion but was often a peak due to a stable fragment ion. The other peaks in the isobutane spectra are usually caused by these same stable ions. An even softer ionization gas, ammonia, with a proton affinity of 205.0 kcal/mole was tried in order to reduce fragmentation further. The ammonia spectra however, also contained many fragmenta- tion peaks. The appearance of the spectra taken using ammonia as a reagent gas, like some other chemical ionization gases, was very 52 100 - El 50-1 1* L Li L- 4 .- °"‘l|"_l'i I I LI I I I I I I I I I 10°“ Methane .J 50-- 0 ll- 1L L us 4L I I I I I I I I I I I I I 10° " lsobutane 50-1 1 o . _L k .L I I I I I I I I I I I I 10°“ Ammonia so— 0 I L l L ALI-J. j I I I r I I I I I I I I ‘°°" Methanol so-l 0 I I I F I I r I I I I LT 4"“!— ‘°°'l Water .l 50- ° I‘IIrTIIIFII‘ILT’Ll-l— 60 80 100 120 140 160 180 200 220 240 260 280 300 320 34-0 Figure 3.5 Mass Spectra of Phosmet 53 100 50 100 50 100 50 100 50 100 50 100 50 El L L I I I* I I I I I I I I ‘ Methane _ L L L L l l - L F—JH I I I I I I I I I I I ‘ lsobutane .111 1..-: A A l J .2; .4 1L 1 J k L A I I’ I I’ I I II I l’ I ' Amnmnm —i and A L A; l L J2 LLL L L A; 1 I* I II I I *I* I I II I I 1 Methanol II I I *I I I“ ‘I 4] II #II I '1 Wate r —i It A . . . , ‘I I I’ I* I I’ II I II I I so so 100 120 140 160 180 200 220 240 260 280 300 Flgure 3.6 Mass Spectra of Phorate 54 dependent on the source pressure. The choice of reagent gas pressure in the source can determine whether the base peak is the (M+H)"‘ ion or a fragment ion. At a source chamber pressure of 0.8 x 10'5 torr, the ammonia CI spectrum of malathion, shown in Figure 3.7, resembles its electron impact spectrum with fragment peaks at m/z 173, 158, 127, 125 and 93 predominating and little (M+H)"‘ ion at either m/z 331 or 348. At this pressure there is not enough reagent gas to create a true chemical ionization environment. The spectrum observed is really an electron impact spectrum of a mixture of ammonia and malathion. At a pressure of 3.8 x 10"5 torr the spectrum begins to resemble a methane CI spectrum. Fragments are still prevalent but there is much more (M+H)"’ ion being formed. As the pressure in the source chamber is increased to 1.2 x 10"“ torr the fragment ions and the (M+H)"’ ions begin to decrease and a (M+NH3)+ ion begins to increase until it becomes the base peak in the normalized spectrum. The change in the spectrum is caused by a change in the proton affinity of ammonia which, like other polar reagent gases, is dependent on the pressure in the source. Polar reagents produce solvated protons, or clusters, of the type (B)nH+ with n as large as three or four. The 205.0 kcal/mole value for ammonia's proton affinity is for a monosolvated proton. The proton affinity for each Cluster size is different and therefore the spectra are different. But, even though ammonia is known as one of the softest ionization gases, the protonation of the organophosphorous molecules by the ammonia clusters was still too exothermic and numer- ous fragment ions were formed. Methanol was also tried as a chemical ionization reagent. H.J.Stan (43) reported that methanol produced only (M+H)+ ions. He therefore made no further use of this reagent system because it did 55 10°“ 0.8 x 10—5 torr 50---I 10°“ 3.8 x 10—5 torr 50 6.0 x 10“5 torr 100-! 50- 1°°'1 1.2 x 10"Jr torr 50" .1 0.1 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 Figure 3.7 Ammonia Chemical ionization spectra of Malathion 56 not give any fragments which he, like most other researchers, desired. Methanol therefore seemed to be the ideal gas for TQMSIuse and, in practice, this was born out. The spectra for almost all of the organo- phosphorous pesticides tested contain only one peak, that representa- tive of the (M+H)"' ions. That methanol should be such a tender ion- ization agent seemed at first unusual because the proton affinity of 184.9 kcal/mole is between that of methane and isobutane. Methanol however, being polar, can cluster and therefore can have a much different proton affinity. Figure 3.8 shows the spectrum of just methanol at chemical ionization pressure. The ions at m/z 33, 65, 97 represent ((MeOH)n.I-H)+ ions. Some researchers add other gases along with the methanol to prevent the clusters from forming, but for organophosphorous esters these protonated clusters are much better reagents than single methanol ions. The clusters seem to be able to give up a proton very gently. Water, because it can also hydrogen bond and is known to cluster, was also tested as a Chemical ionization reagent. The resul- ting spectra were, like the methanol spectra, very simple. The base peak was again from the (M-I-Hi)+ ions, but this peak was not the only peak observed. Phosmet, parathion, phorate and malathion all had small peaks from fragment ions as well. These fragment ions are formed by simple cleavages or the loss of small stable neutrals and are very indicative of the sample compound. Water was not chosen as the optimum gas because it produced a small fragment ion current and it was not as easy to use as methanol. At room temperature, water's vapor pressure is not high enough to give a pressure that is easily regulated. The highest pressure reached was not optimum and fluc- tuated slightly giving pulsating peak intensities. Attempts to heat 57 [(C3308)2+H]+ 100 ‘7 I+m* l [(CHBOH)3+H]+ 50 -' 0 J J i L 7" 'l ' l‘iT '71 I 'l ' l' I ‘17T l 20 40 60 80 100 120 Figure 3.8 Background Spectrun of Methanol Reagent Gas 58 the water reservoir and inlet system in order to raise the pressure proved to be a lot of trouble and were eventually abandoned. Methanol was chosen as the chemical ionization reagent. The resulting spectra have, in general, one peak representative of the (M+H)"’ ion. As is demonstrated in the next chapter, the (M+H)"’ ions formed can be fragmented in the collision cell to give good daughter spectra. Methanol is also readily available and has a vapor pressure at room temperature that is high enough to allow easy regulation. 59 (hapter 4 Detection of Organophosphorus Pesticides Intmductian Looking at the daughter spectra of the pseudo-molecular ions of the organophosphorus pesticides revealed several ions and fragmentation mechanisms indicative of the parent compounds. The fragmentations, which were discussed in chapter two, can be employed by a triple quadrupole mass spectrometer to screen for organophos- phorus pesticides. A small number of parent scans and neutral loss scans can be used to detect almost all the 38 compounds under study. The goal is to find the smallest number of scans that need to be used to detect all the possible organophosphorus pesticides. ‘Table 4:1 lists twelve scans that could be used to detect all the pesticides listed. Other organophosphorus compounds, unknown pesticide metabo- lites and decomposition products could also be detected with these same scans“ Several of the listed scans have been tested to determine how well they work. W The first detection study was done with seven phosphoro- thioates, four with methoxy R groups and three with ethoxy. Isobutane was used as the CI reagent, because the benefits of methanol had not yet been discovered. Six of the pesticides, dialifor, phosmet, azin- phos ethyl, dimethoate, thiometon and disulfoton, were obtained from 60 Table 4.1 Scans Used to Detect Organophosphorus Pesticides SCAN MASS .IIEE .SELEQIED EESIIQIDE§.EQUND parent 109 pirimiphos methyl fentrothion parent 125 heptanOphos methyl parathion quinalphos etrimphos methyl dursban diazinon oxygen analog dimethoate of dimethoate parent 127 heptanophos dicrotophos naled phosphamidon tetrachlorvinphos malathion chlorfenvinphos oxygen analog of malathion parent 153 diazinon dioxathion ethion parent 155 chlorfenvinphos oxygen analog of dimethoate parent 169 diazinon neutral 137 pirimiphos quinalphos loss neutral 142 oxygen analog oxygen analog pirimiphos loss of demeton of phosmet methyl dimethoate neutral 152 dursban pirimiphos loss neutral 158 phosmet azinphos thiometon loss neutral 170 demeton-S demeton-0 quinalphos oxygen analog of phorate neutral 186 disulfoton dialifor dioxathion loss azinphos ethyl ethion phorate 61 Chem Service, Inc. and were used without further purification. The malathion used was a commercial pesticide formula, ORTHO Malathion 50, obtained from a local garden store. Instrument grade isobutane and argon were obtained from Mathison. The solid and liquid pesticide samples were mixed together and diluted to a concentration of about 2 micrograms per microliter with ether and the ORTHO formula. One or two microliter samples were placed in a capillary melting tube and inserted into the mass spectrometer on a solids probe. Spectra were taken on an Extra-Nuclear ELQ-400-3 triple quadrupole mass spectrome- ter and data system. The spectra were obtained using neutral loss and parent scans using argon as the collision gas at a pressure of 1.8 x 10‘” torr. W A normal mass spectrum of the sample mixture was obtained by scanning the first quadrupole and keeping the other two quadrupoles in the RF only mode. The base peak, at m/z 331, of the spectrum was due to the pseudo-molecular ion of malathion. Most of the abundant ions present were fragments of malathion. This was expected since there was 100 times more malathion in the mixture than the other pesticides. The pseudo-molecular ions of the other species in the mix were almost visible in the multitude of peaks with less than 1% abun- dance. Most of the other ions from the "inert ingredients", emulsi- fiers and xylene-range solvents present in the ORTHO sample do not show up in the spectrum because their masses are below 100. One can not scan below m/z 100 when using isobutane as a reagent gas because its ion at m/z 97 is much larger than any of the sample ion 62 intensities. The other pesticides present in the mixture can be detected however, by using neutral loss scans. Neutral loss scans of 158 and 186 mass units were used to obtain the spectra shown in Figures 4.1 and 4.2 These two scans should be able to detect any dithioate. Disulfoton, azinphos ethyl and dialifor were detected at m/z 275, 346 and 394-397 respectively using the neutral loss scan of 186 mass units. These three compounds were expected to show up in this scan because they are all ethoxydithioates whose daughter ion spectra indicated a tendency to lose a 186 mass neutral. The peak at m/z 318 however was not expected. This peak is from phosmet, a methoxydithioate, which should lose 158, not 186 mass units. The 186 mass loss is from a double elimination fragmentation. The base peak in the daughter ion spectrum of phosmet is at m/z 160, from the breaking of the Y-Z bond and the charge staying with the Z group as discussed in Chapter 2. The daughter ion spectrum of the m/z 160 ion shows ions at mass 132, 104 and 76, indicating successive losses of 28 mass units. Therefore it can be assumed that the loss of 186 mass units from phosmet is the total from a loss of 158, yielding m/z 160, and then a loss of 28 to give m/z 132. This double neutral loss could either be caused by a too energetic collision or multiple collisions. The multiple collisions could possibly be corrected by using a lower collision gas pressure but if it is lowered too far, then not enough of the molecular ion is fragmented and sensitivity is lost. ‘The 20 volt collision energy was determined during the acquisi- tion of daughter spectra and was chosen because it gave the best abundance of the ion representing the desired neutral losses. The energy is therefore fixed as the best choice and for consistency 63 Intensity 100 80 60 40 20 .J Phosmet 4 q Thiometon l L II'I'IIII'IIT'III 200 220 240 250 280 300 320 340 360 380 M/Z Figure 4.1 158 Neutral Loss 64 Intensity 100 80 60 40 20 _. 1 Aflnphos Ethyl 4 DiaHfor Phosmet 1 Disulfoton f 1 I l T 1' 1 I I I I I FT I l I I 1 200 220 240 260 280 300 320 340 360 380 400 M/Z Figure 4.2 186 Neutral Loss 65 should not be adjusted for each compound. The other compound in the mixture, dimethoate was not detected using neutral loss scans, as expected. Even though it is a methoxydithioate it does not fragment by breaking the Y-Z bond. It undergoes the unexplained rearrangement discussed in chapter two, and therefore does not lose 158 mass units. A parent scan obtained by setting the third quadrupole to pass only ions with m/z 125 and scanning the first quadrupole is shown in Figure “.3 . This scan should detect any of P1 ions from the dimethoxy dithioates and those diethoxy dithioate P1 ions that undergo a Quale rearrangement. This scan did detect dimethoate at m/z 230 as well as thiometon at m/z 2H6, phosmet at m/z 318, and two large malathion peaks at m/z 285 and m/z 331. There are also several other unexplained peaks, probably due to the fragmentation in the source of several of the compounds. Again this demonstrates the importance of creating only molecular ions in the source. The success of these sample scans showed the feasibility of this screening method and prompted further experiments. After examining different CI reagent gasses and settling on methanol, a second detection experiment was carried out using a broader range of compounds . W11 Ten compounds were chosen that represent dithioates and thioates with both methoxy and ethoxy R groups. The ten compounds chosen are: fentrothion, dursban, quinalphos, parathion, etrimphos, malathion, dimethoate, phorate, phosmet and dialifor. In an attempt to eliminate the problems of fragmentation in the source, methanol was 66 Intensity 100 80 60 40 20 .4 Malathion q .+ ' M l th’ ‘ Dlmethoote no Ion .1 _ Thiometon Phosmet u I I r T I r 'I I I l I u I , I r I 200 220 240 260 280 300 320 340 360 380 400 M/Z Figure 4.3 125 Parent Soon 67 used as the CI reagent gas. The methanol was injected with a syringe into an evacuated expansion bulb and its vapors were then leaked into the source through a needle valve. The pressure in the source was adjusted until the methanol cluster ions at m/z 65 and 97 were maxi- mized and the ions at m/z 33 and 97 were approximately of equal intensity; Measured amounts of the ten pesticides were mixed in a xylene solvent and diluted to a concentration of about two micrograms per milliliter. The sample was introduced into the mass spectrometer with a solids probe into which was fitted a capillary tube containing one to two microliter samples of the mixture. The solids probe was slowly heated to 150 degrees centigrade, allowing the pesticides to volatilize. Six scans were used to try to detect the pesticides. As in the first experiment, neutral losses of 158 and 186 mass units were used to detect the dithioates present. A 142 mass unit neutral loss was used to detect the methoxy thioates present and a 152 mass unit neutral loss was tried as a way to detect the diethoxy thioates. Two parent scans were included. Scans for the parents of ions with m/z 125 and 127 were used to detect fentrothion and malathion because their ions due to the above neutral losses were small. Two different automated data collection software programs *were used to take data. Software designed for chromatography has been implemented on the Extra Nuclear ELQ-400—3 mass spectrometer that continuously repeats a programed set of scans. In this case the instrument was instructed to do a scan for a neutral loss of 1H2, then successive scans for neutral losses of 152, 158 and 186, then do scans for all the parents of m/z 125 and 127. This sequence was repeated every two seconds for 600 seconds or until stopped by the operator. 68 The other program used utilizes multiple reaction monitor- ing to detect compounds as they elute off a gas chromatography column, or in this case, off the probe. Up to eight different specific parent- daughter ion pairs are entered into the program which, when run, continuously cycles through the pairs, setting quadrupoles one and three to the correct values and detecting any ions that get through in a specified amount of time. Selected ion monitoring experiments may also be done by setting the first quadrupole to pass only the ion of interest and putting quadrupole three into the RF only mode. Adduct ions, caused by the addition of mass during the collision process, may also be monitored by setting the daughter to a higher mass than the parent. B J! I D' . II Nine of the ten pesticides were detected using the first program. Because of differences in volatility, not all the pesticides were ever detected in one set of scans. Dimethoate could be detected as soon as the probe was inserted. Other compounds were detected as the probe slowly heated to 150 degrees where Dialifor could be detec- ted. Figure "w” shows averaged composite spectra for the scans used. The only compound not detected was Parathion. llzhas a small peak due to a loss of 152 mass units in its EI spectrum, but this ion is not formed by collision. The prominent ions in the Para- thion collision spectrum are due to losses of 28 mass units. This neutral loss was not used to detect the pesticide because loss of 28 is.a very common neutral loss and thus not indicative of just the pesticide. The other neutral losses used in this study are much more 69 142 Neutral Loss 100 H g 'l Dimethoate "8 o 50 - H 5 - O‘l¥l‘l‘l’l'l'l'l'l'l 152 Neutral Loss 100 — . >‘ _ Quinalphos +1 ":5 g 50 -d Dursban *5 . . - Fentrotbion L . l A 0 r I I T I I I I I l I r I I I l' I l 158 Neutral Loss 100 — Phosmet 3. .. '6 S 50 -l +5 Malathion . — ‘ L Malathion 0 ‘l t l I r I l I I I r IL I' I I I l I 1 186 Neutral Loss 100 - 3:" _ Phorate 2 50 ._ Dialifor 3 _g_: - 0 l r I I ii I l’ I 17f r T l I I I l I 125 Parent Scan 100 -l . Dimethoate >. _ nub-J "é o 50 - 4.! E q 0 ll L l L l I ‘ I ' I ' 1 ‘ l ' I ' l ' I ' I 200 240 280 320 360 400 M/Z Figure 4.“ Spectra Obtained Using Scanning Techniques 70 uncommon and therefore should not suffer from interferences from sample matrix ions. Dimethoate, which was not detected using the neutral loss scans in the first experiment, was detected in the HQ neutral loss scan. Dimethoate, a dimethoxy dithioate, should not be detected by a scan targeted for dimethoxy thioates but it is because the base peak at m/z 88 in its CAD spectrum (the unexplained rearrangement ion) just happens to be 1112 mass units less than the m/z 230 pseudo-molecular ion. This same sort of situation allows for the detection of malathion using an m/z 127 parent scan. Malathion can be detected with the expected 158 mass unit neutral loss, but because the m/z 173 peak that is needed for this scan to work is very small, malathion is not always picked up. Since the base peak in the CAD spectrum of the pseudo-molecular ion is at m/z 127, it can be detected much more easily, with potentially better sensitivity, using the m/z 127 parent scan. Using this scan does not increase the total number of scans used since the m/z 127 parent scan is also used to detect the dimeth- oxy phosphates. Those scans that accidentally detected pesticides point out an unavoidable problem and an important consideration. Using a reagent gas that makes only pseudo-molecular ions does help eliminate ambiguity in compound identification, but does not get rid of all interferences. The two pieces of information in a neutral loss or parent scan, the m/z of the parent and the neutral or daughter mass can be used to identify a known compound but not an unknown compound. If for example, the 1112 neutral loss scan yields a peak at an m/z that does not correspond to a known pesticide, the unknown compound does 71 not have to be a dimethoxy thioate, or even an organophosphorus com- pound. More information is needed to make an identification, such as a complete daughter scan of the detected ion. By using unusual neutral loss and uncommon parent scans, chances of detecting only the desired compounds are better, but not assured. Any scanning technique which, by design looks for anything, can detect.anything. 'This.is good for metabolite or decomposition studies in which the metabolized species is not known, because a scanning technique will be able to detect the unknown compound. For targeted analysis however, in which only specific compounds need to be detected, a scanning technique that could potentially produce a peak at an unexpected mass could cause problems. Reaction monitoring of specific known ion pairs can eliminate this problem but then only the known compound can be detected. Therefore the goals of the experiment should be considered when choosing the type of scan used. The ion current chronograms from the multiple reaction monitoring experiments are shown in Figure “AL All eight of the pesticides were detected. The peaks are broad because they represent heating profiles and not GC peaks. The temperature was initially allowed to rise to 100 degrees, but after approximately 280 seconds it was raised to 150 degrees in order to vaporize dialifor. The number in the upper left hand corner of each plot represents the ion current at the top of the peak in ADC units. The differences in the numbers is due in part to the different concentrations of the various com- pounds in the mixture, but also due to the relative abundances of the selected ions in the parent ion spectrum. The fragmentation efficien- cy and therefore the daughter ion abundance are dependent on the collision energy, collision gas species and pressure, and the mass of 72 230.0 -> 88.0 45°41 Dimethoate O. Trrlrlll'lrrIl’erIIIlrfir 261.0 —> 75.0 44017. 3 /\ Phorate 0' l—‘I’rTrIrTIIIIIIIITTIFIIITTI'jIIVIIIITrITIIrTIIIlI 278.0 -> 125.0 .3728. Fentrothion 0°jIITITITTjrjfirIIIIIIIIIIIIIITfirT'IIIIlIFIIlITiTII 299.0 -—> 147.0 11755. a Quinalphos 0' ‘IIIIIIIITIIIIIIIIII'IIIIIIIrIIIr—II'I[ITIIIITfi 318.0 —-> 160.0 100583- a Phosmet / O.jIIrT'IIrIlIIIITIrrI'IIII'IIII'IIIIIIIIIIIIII'I 331.0 -> 173.0 450' ”I Malathion 0c IIIIlIIITITrIIleTTIIIII'I 350.0 —> 198.0 5990. Dursban o.jIrIT'IIIIIIIIIIFIIIIIIIITTIIIIIIIIIIIIT—IrITIl—T' 394.0 -> 208.0 18035. Dial ifor 09 rjfiTlIIII'IIjrfirIrrIIIfiITIIIrIT‘T’leII—I‘IfiTITV 0 50 100 150 200 250 300 350 400 450 Time (sec) Figure 4.5 Multiple Reaction Monitoring Results 73 the parent and daughter. Each reaction has a different set of condi- tions for maximum efficiency and, when monitoring more than one reac- tion at a time, compromises must be made that adversely effect some of the reactions monitored. The malathion ion current is the smallest however, because the m/z 173 ion is not a very abundant ion in the CAD spectrum of the m/z 331 parent. Reaction monitoring, because it does not have to scan the mass filters, is a much more sensitive technique than the method involving the neutral loss and parent scans. The ion currents detec- ted using the multiple reaction monitoring programs were several times larger than those from the scanning programs which were near the detection limit of the instrument. A 100 fold dilution of the mixture was made and analyzed again using the same monitoring program. Approximately twolnanograms of each compound was present in the sample analyzed. The ion current chronograms for the dilute mix are shown in Figure MIL This time only seven of the eight compounds were detec- ted. The peaks in the dialifor chromatogranlare just noise spikes with three ADC units of intensity. It is not known why the dialifor was not detected in the dilute solution. Malathion was still detect- able, even though the signal to noise was not very good. The sensitivity of this technique could be improved fur- ther by looking for only one compound and tuning the instrument to favor the reaction of interest. Detection limits in the picogram to femtogram range should be attainable. 74 230.0 —-> 88.0 ”:3 1k 0 ITII ITIITITIII TTIII I “I T I T I 261.0 —> 75.0 12. AA“ AA 0. I'IIT'rrTTTIrT OO—IIIIIIIIII'IIIIIIIIIII 278.0 -> 125.0 299.0 -> 147.0 -> 106.3 A- 0. [TIIIIAIIITII IITIIIIITI 318.0 —> 160.0 :3 A ojIIII'IjII'IIIIITITIrI 331.0 —> 173.0 I ITT T T I l I T ITI— O A A l A o I I I I l I T r I I I T I I ' I I T T l I I l I I I I I I T I I 350.0 —> 198.0 142. a k 0. I I I I I I T T j I I I T I I l T I I I ‘ I T TT I I I I I T I TI 394.0 --> 208.0 3. a A l 1 l lll l A O. I I I I I l r r I I I I I I l T I if I I I I r I I I I I r T I T 0 50 1 00 1 50 200 250 300 Time (sec) Figure “.6 Multiple Reaction Monitoring Results (dilute sample) 75 gonclusion By only looking for one specific collision reaction pro- duct the sensitivity and selectivity of the analysis are enhanced, but only the compound of interest can be detected. Knowing what to look for has the advantage of allowing the instrument to be tuned to favor the production and detection of the reaction components. By monitor- ing several reactions with one set of instrument conditions, sensi- tivity is lost. This could be corrected for by allowing the computer to adjust the variable instrument parameters so that the best settings for each compound are being used when data is being taken. Unknown or unexpected compounds can only be detected using scanning techniques. Neutral loss and parent scans can identify known compounds, but are not as sensitive. Scanning techniques can detect an unlimited number of compounds at once with a common functionality, but the absolute identity of some may require confirmatory daughter scans. Both methods are capable of routinely detecting many pesticides on a short time scale. 76 REFERENCES References 1. MelnikovaN”"Chemistry of Pesticides",Springer-Verlag, New York (1971) 2. Emsley,J. and HallJL, "The Chemistry of Phosphorus", Halsted Press, New York (1976) pg.494-509 3. Berza,Mn, "Pestmanagement with Insect Sex Attractants" ACS Sympo- sium Series #23, Am.Chem.Soc., Washington D.C. (1976) 11. Ware,G.W., "Pesticides; Theory and Application", W.H.Freeman and Co., San Francisco (1978) 5. Leomhardt,B.A. and Berza,M.,"Insect Peramone Technology: Chemistry and Application", ACS Symposium Series #190, Am.Chem.Soc.,Washing- ton D.C. (1982) 6. Harris,CJL, Persistence and Behavior of Soil Insecticides,in "Pesticides in the Soil", International Symposium on Pesticides in the Soil, MSU Press (1970) 7. Harris,C.R., J.of Economic Entomology, 92 (6),“438 (1969) 8. Fest,C. and Schmidt,K“L, "The Chemistry of Organophosphorus Pest- icides", Springer-Verlag, New York (1973) 9. The Merck Index, M.Windholz (ed.) 9th edition, Merck and Co., Rahway'NuL (1976) 10. Spencer,w.F., Distribution of Pesticides Between Soil, Water and Air, in "Pesticides in the Soil",International Symposium on Pest- icides in the Soil, MSU Press (1970) 77 11. 12. 130 11-1. 15. 16. 17. 18. 19. 20. 21. 22. 23. 211. 25. 26. 27. 28. 29. Seiber etal., in "Pesticide Analytical Methodology",Zweg,H. (ed.) ACS Symposium Series #136, Am.Chem.Soc. Washington D.C. (1979) pg.179 Ambrus,A. etal., J.Assoc.0ff.Anal.Chem.§_fl (3), 733 (1981) Halmann,M., "Analytical Chemistry of Phosphorus Compounds", Wiley Interscience Press, New York (1979) Saner,W.A. andGilbert,J.,J.of Liquid Chromatography 3 (11), 1753 (1980) LeBel etal., J.Assoc.0ff.Anal.Chem. .62 (2), 2141 (1979) Thomas,T.C. and Seiber,J.N., Bulletin of Environmental Contamination and Toxicology, 12 (1), 17 (19711) Woodrow,J.E. and Seiber,J.N., Anal.Chem. 50 (8), 1229 (1978) Bidleman,T.F. and Olney,C.E., Bulletin of Environmental Contamina- tion and Toxicology, .11 (5), 11112 (1971-1) Lewis,R.G. etal., Anal.Chem. £9 (12), 1668 (1977) Lewis,R.G. and Jackson,M.D., Anal.Chem. 53 (3) 592 (1982) Lewis,R.G. and McLeod,K.E., Ana1.Chem. 5}; (2) 310 (1982) Otsuki,A. and Takaku,T., Anal.Chem. 51 (7), 833 (1979) McCully,K.A., J. Assoc.0ff.Anal.Chem. .63 (2), 283 (1980) Sherma,J. and Zweig,G., Anal.Chem. 55 (5) 57R (1983) Smart,N.A. etal. Analyst 103, 770 (1978) Ministry of Agriculture, Fisheries and Food, Analyst 102,858(1977) Shafik,M.T. and Enos,H.F., J.Agri.Food Chem. 11,1186 (19611) Lee,Y.W. and Westcott,N.D.,J.Assoc.0ff.Anal.Chem. 62 (11), 782 (1979) Biros,F.J., "Pesticide Identification at the Residue Level", Advances in Chemistry Series #1011, Am.Chem.Soc., Washington D.C., (1971) 78 30. Faust,S.D., "Fate of Oganic Pesticides in the Aquatic Environment", 31. 32. 33. 311. 35. 36. 37. 38. 39. 110. 111. 112. 113. L111. 115. 116. 117. Advances in Chemistry Series #111, Am.Chem.Soc., Washington D.C. (1972) McLarrerty,F.W., "Interpretation of Mass Spectra", 3rd edition, University Science Books, Mill Valley, Calf. (1980) Budzikiewicz etal., "Mass Spectrometry of Organic Compounds", Holden Day, San Francisco. (1968) Silverstein,R.M.,etal."Spectrometric Identification of Organic Compounds", 3rd edition, John Wiley and Sons,New York, (197M) Domico,J.N., J.Assoc.Off.Anal.Chem. 59, 1027 (1966) Jorg,J. etal., Monatshefte fur Chemie 91, 1065 (1966) Stan,H.J. etal. Fresenius Z. Anal Chem 281, 1065 (1977) Desmarchelier,J.M. etal., Residue Reviews 63,77 Safe,S. and Hutzinger,O., "Mass Spectrometry of Pesticides and Pollutants", CRC Press, Cleveland Ohio, (1973) Cooks,R.G. and Gemard,A.F., J.Chem.Soc.B, 1327 (1968) Biros,F.J., Ross,R.T., 18th Conf.on Mass Spec, Allied Topics, San Francisco, 1970, paper #03 Quale,A., The Mass Spectra of Some Organic Phosphates, in "Adv.Mass Spec.", vol 1, Waldron,J.D.(ed.), Pergamon Press, London, (1959) 98-365 Holmstead and Cassida, J.Assoc.Off.Ana1.Chem.,51, 1050 (19711) Stan,H.J., Fresenius Z.Anal Chem. 281, 1011 (19711) Busch,K.L. etal., Applied Spectroscopy, 32 (it) 388 (1978) Daughtery,R.C.,Wander,J.D.,Biomed Mass Spec., 1(9), 1101 (1980) Rankin, P.C., J.Assoc.0ff.Anal.Chem. 55. (6), 13110 (19714) Stan,H.J., Kellner,G., Bio.Med.Mass Spec.,9_,(11), 1183, (1982) 79 APPEIDH Intensity 100 80 60 4O 20 Aspon daughter spectrum of the MH+ ion 11 1 1 l ' I I ' I . I I T I I I I ‘1 so 100 140 180 220 260 M/Z 021150 021450 \s _S/OCZH5 / \OCZHS Intensity Azinphos daughter spectrum of the MH+ ion 100 .1 80‘- 60 -1 4o- 60 100 140 180 220 260 M/Z A2 300 Intensity 100 80 a) O .5 O 20 Azinphos Ethyl daughter spectrum of the MH+ ion i .1 III [llih l I I L ‘I'I'T'l‘ 'I'I'I'I'I' IIII so 100 140 180 220 260 300 340 M/Z 92H50\§ oszo/ Intensity Chlorfenvinphos daughter spectrum of the MH+ ion 100 - 80 - 60‘- 40-1 20 - O Ill'IlTII'FITI’TjI'IlI[I'T‘TrI 60 100 140 180 220 260 300 - 340 M/Z CL CZHSO\SID P—O --— CL / CszO CL—CH A4 D e m eto n — 0 daughter spectrum of the MH+ ion 100 -I 80 -1 60 -I Intensity 4o— 20 - 0 rT ltri‘rITrITrTIIl 60 100 140 180 220 260 M/Z cszo \fi /P-O—CH2—CH2- $112—ch3 CZHsO A5 Intensity Demeton—S daughter spectrum of the MH+ ion 100 - 80- 60-1 40- 20- O‘II’T‘I‘IfiI'I'I'I‘I so 100 140 130 220 260 M/Z 0211503 /P—s—CH2—CH2—s—CHz-c H3 cszo A6 Intensity Dialifor daughter spectrum of the MH+ ion 100 '1 80-1 50-4 40- 20- .: l_ O i .424 i - 3 ii A 22, 'I‘IIIII‘I‘I'TTT‘T‘I'I'I‘I’I'I‘I'I 60 100 140 180 220 260 300 340 380 M/Z 0me / \ C we S-in-Nf I) CL-CH 0 A7 Intensity Diazinon daughter spectrum of the MH+ ion 100 -I 80-1 60 - 401- - II I, o I ' I l I I T l I IT I I I I I I I I 'fi 60 100 140 180 220 260 300 M/Z A8 Intensity Dicrotophos daughter spectrum of the MH+ ion 100 '- 80-4 BO-I 40— 20--l O‘I'I'I‘I'I’I‘I'I‘Trl so 100 140 180 220 260 M/Z A9 Intensity Dimethoate daughter spectra of the MH+ ion 100 -1 80 -‘ 60‘- 40 -1 20 - T I I T r f r If I I I t I I 60 100 140 180 220 M/Z CH30\§ o \ CHSO/ 8"CH2— ‘CH3 (":3 I—Z A10 Intensity Oxygen Analog of Dimethoate daughter spectrum of the MH+ ion 100 - 30- 60- 4o- 20 -‘ 60 100 140 180 220 -N—CH3 A11 Intensity Dioxathion daughter spectrum of the MH+ ion 100 -I 80 - 60 -I 40 -* 20 -1 O III'I Ill'r'llIrIITIl 'rrIlIIIII'TlIII‘Il 60 140 220 300 380 460 M/Z 02H 0 9 5 \P—S CZHSO/ j 02H50\fi p—s CZHSO/ Intensity Disulfoton duaghter spectrum of the MH+ ion 100 -I 80 -1 60- 40 - 0 I I I I I I ' I I I 1* I . I . I . I . I . I . I so 100 140 130 220 260 300 M/Z cam; c H o/ *S-CHz-CHZ-S—CHz—CHB E! 55 A13 Intensity 100 80 60 4o 20 Dursban daughter spectrum of the MH+ ion .1 L l 1 I'I‘I‘I'IIITT‘I‘I‘I ITI'Ifi 60 100 140 180 220 260 300 340 11/2 CL cszo 8 ll \P—O—/ \ CL C H(D 2 5 CL A14 100 80 60 Intensity 40 20 Methyl Dursban daughter spectrum of the MH+ ion --l A A l I l 1 J I r T I T r I I l I l I rI rI l I l I l I I I I F] 60 100 140 180 220 260 300 340 11/2 ergo CH3O A15 xi p Intensity Oxygen Analog of Dursban daughter spectrum of the MH+ ion 100 -I so- so— 40-- 20... o I I I l I ITI ' I l I T F I I I I r I I I I T 1 l I I 60 100 140 180 220 260 300 340 M/Z cszo / N— CHO 25 CL A16 Intensity Ethion daughter spectrum of the MH+ ion 100 - ao- sod 40- 20a ‘ l J 0 ‘I I I 'I I I' I I ' I IAFI ' I 'I ' I' I ' I' Ii' I' I 60 100 140 180 220 260 300 340 380 M/Z 02H50\fi fi OC H /P—S-CH2—S-—P< 2 5 A17 Intensity Etrlmfos daughter spectrum of the MH+ ion 100 - 80 -1 60 -1 4o- L l 7 I ' I ' I I ' I '7 I ' I ' I ' I I I I 60 100 140 180 220 260 300 M/Z Intensity Fe ntroth l o n daughter spectrum of the MH+ ion 100 -i 80 -1 60 -‘ 40‘- 204 O‘H‘th l L AL L LA I I k ' I ‘ I ' I ' I ‘ I I I ' I ' I ' l ‘ I ‘ I 60 100 140 180 220 260 300 M/Z <3H51\e CH3O org A19 Intensity Heptenophos daughter spectrum of the MH+ ion 100 -[ so- 60-- 4o- 20.— 0 I I I l I I I [ fl I I I I I 1' 1 I ' I’ 1 60 100 140 180 220 260 M/Z CH30 0 ll HD—Oflifi \ CL CH3O A20 Intensity Malathion daughter spectrum of the MH+ ion 100 - 80- so- 40-- 20-1 4L 4_.4c J L L_ 1 J o‘rflA—M‘L‘I'I‘I'F'T'I‘I'I'IFI I 60 100 140 180 220 260 300 340 M/Z A21 Intensity Oxygen Analog of Malathion daughter spectrum of the MH+ ion 100 - 80 - 60 -+ 0 L A A J L .1 I I i I I I T T l I l I l I I I I I l I j 60 100 140 180 220 260 300 M/Z CHsO O )5\ 9 8‘ H—C—O—Csz CHSO H2_(é—O _CzH5 A22 Intensity Noled daughter spectrum of the MH+ ion 100 - 80 -< 60 -‘ 40-4 20 -i O IlIlleIIIIII'IIIlrrIrIIWITIrII 60 100 140 180 220 260 300 340 380 M/Z A23 Intensity Oxydenneton Methyl daughter spectrum of the MH+ ion 100 '- 80- 60- 404 20- Ortrljl'l‘ltrlrl'lTfi so 100 140 180 220 260 M/Z c H30\g l CH3 0/ \S—‘CHZ‘CHZ‘é—CH2-‘C H3 A24 Intensity Porothion daughter spectrum of the MH+ ion 100 -1 80-1 60-1 40- 20-l 60 100 140 180 220 260 C 2H 50\§| Cszo A25 Intensity Methyl Parathion daughter spectrum of MH+ ion 100 — 80- 60- 40-4 ZO-l A26 Intensity Phorate daughter spectrum of the MH+ ion 100 - 804- 60 -1 4o- 20 -« 0 L k l I I I 1 I I I I I I I I ' I I r I fin I r r f I— ] Ifi 60 100 140 180 220 260 300 M/Z CZHSO\E s CH CH CH Cszo/ 2 2 3 A27 Intensity 100 ”v Oxygen Analog of Phorate daughter spectrum of the MH+ ion aa- 50-- 4o- 20 -‘ o I I j l' f I I I I— I I I I I I '1 I ' 60 100 140 180 220 M/Z CZH50\E 2 5 A28 Phosmet daughter spectrum of the MH+ ion 100 -I 80-1 60--l Intensity 4o- 20- 0 . L Ij r I I I f I I I I I I I I I T r I I I I I I ‘ I I 60 100 140 180 220 260 300 340 M/Z s CH3O\.. —s —-—CH O/ 2 “(If CH3O A29 Intensity Oxygen Analog of Phosmet daughter spectrum of the MH+ ion 100 - 80 - 60 -' 4o- 20 -I I I I I I I I l T j 1 I r j I I I r 60 100 140 180 220 M/Z =0 CH o\§ CHsO/ \—-N\S—CH2 CD-‘T-‘CU/£ \O ' A30 I 260 I T I I I 300 100 - 80- 60— Intensity l 40-- 20- phosphanfidon daughter spectra of the MH+ ion T I r l r r I 100 C H30\ OPE 140 A31 ITI' 180 M/Z ' I 220 u C.”- P—o—C=C— —N CH3 I I r r I 260 \Csz /CzH5 300 Intensity Pirimiphos daughter spectrum of the MH+ ion 100 -l (02H 5)?_ Intensity 100 '1 80- 60--l 40-1 20-I Pirirniphos Methyl daughter spectrum of the MH+ ion A33 Quinalphos daughter spectrum of the MH+ ion 100 '- 80 '- >‘ 60 4 4.: 75 c .. o 4.0 E 40 - 20 -I 0 L 1 L I I I I I I I I I I I I I I I 60 100 140 180 220 260 300 M/Z CZH50\{L|1 —O / Cszo/ fl” A34 Intensity Temphos daughter spectrum of the MH+ ion 100 - ao-l 60- 40- 20-4 0‘ l ll III'I'IIIII'IIIFIII lIIIlI'I'I'ITj'I IlIlI 60 100 140 180 220 260 300 340 380 420 460 M/Z A35 Intensity Tetrachlorvinphos daughter spectrum of the MH+ ion 100 -I aa- 60-- 4o— 20 -4 O I'rTITIrfilITIlI‘lI'I‘fIrITTIIT'rlII 80 100 140 180 220 260 300 340 380 M/Z CL CH 0 $3 3 \P—O-C CL / ll 0 — CH3 CL CH \CL A36 Intensity 100 - 80 -‘ Thiometon daughter spectrum of the MH+ ion 60 l 1 ll Irj‘rrlITrIIIer'leI 100 140 180 220 260 M/Z CH30\E \ CH 30/ s -C Hz—C Hé-s—C Hz-C H3 A37 "Illlllllllllllllllllllllls