MS and Time of Flight

---

Chapter 6

Determination of an Unknown System Contaminant Using LC/MS/MS 1

1

1

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

2

Brett J. Vanderford , Rebeeca A. Pearson , Robert B. Cody , David J. Rexing , and Shane A. Snyder 1

1

Southern Nevada Water Authority, 243 Lakeshore Road, Las Vegas, NV 89153 JEOL USA Inc., 11 Dearborn Road, Peabody, MA 01960 2

In recent years, liquid chromatography coupled with tandem mass spectrometry (LC/MS/MS) has become an increasingly valuable tool for the study of contaminants in the environment. It has the advantage of providing structural information while simultaneously reducing background interference. The objective of the investigation presented here was to identify an unknown system contaminant using high¬ -resolution accurate mass measurements and LC/MS/MS. In order to accomplish this, a reverse-geometry, double-focusing mass spectrometer equipped with linked scan capability was used. After an accurate mass measurement was performed, an elemental composition determination was carried out to generate a list of potential compounds. Once the list was narrowed down to the most likely molecular formula, linked scan MS/MS was used to provide enough structural information to confirm the chosen formula and identify the most probable constitutional isomer. The unknown contaminant was determined to be N¬ -butylbenzenesulfonamide, a common plasticizer increasingly found in the environment.

96

© 2003 American Chemical Society

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

97

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

Introduction Recent reports have shown that certain contaminants at trace concentrations in surface waters can have dramatic reproductive effects on aquatic organisms (1-3). These compounds are collectively known as endocrine disrupting compounds (EDCs). Pharmaceuticals and personal care products (PPCPs) have also been detected in the aquatic environment and may act as EDCs (4-7). As more environmental contaminants such as these are discovered, the need for sensitive detection and identification methods has also increased. Traditional gas chromatography is of limited value without time-consuming derivatization because many environmental contaminants are polar, have low volatility and are thermally labile. This has led to the increased use of liquid chromatography/mass spectrometry (LC/MS) due to its ability to effectively analyze these types of molecules. Since many of these contaminants are found at extremely low levels in the environment, extraction procedures are often used to concentrate them to detectable levels. However, these methods generally concentrate not only the compounds of interest, but also high levels of unwanted background material such as natural organic matter. This may lead to the misidentification of target compounds due to the increased chance of co-elution with compounds having the same monitored mass. In order to overcome this problem, MS/MS techniques have been developed that effectively separate the compounds of interest from background interferences through the monitoring of precursor/product ion pairs. A second use of MS/MS is for the identification of unknown compounds. Since the analyte undergoes fragmentation to create product ions, useful structural information can be obtained by studying a product ion scan. By calculating the difference in mass between the precursor and product ions, fragment compositions can be determined which lead to structural elucidation. In this investigation, a high-resolution mass spectrometer with MS/MS capability was used. The coupling of high-resolution and MS/MS yielded a powerful combination of structural information and accurate mass measurements. These techniques were utilized in a series of experiments to determine the structure of an unknown system contaminant that was suspected to be causing suppression of the EDCs and PCPPs of interest.

Background During routing analytical work, it was observed that an unknown contaminant was constantly producing an extremely large background when the

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

98 mass spectrometer was operated in full scan mode. This is shown in Figure 1. The signal was consistently at 100% of full scale and ion suppression was suspected. Even when the photomultiplier was attenuated by a factor of 0.25, the signal was readily observable, suggesting a complete saturation of the detector and probable negative effects on detecting any compounds of interest.

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

D0108-13 20ul 4 Mix 250ppb each with MeOH grad. 65-100 55 min. 0.2mL/min Scan: 979 TIC=3729271 Base=100%FS #ions=591 UV Signai:898.4

80-

60% 40-

20- 85.1017 266.2211

447.2997

139.0995 m/z

100

200

300

' Ί »' 400

I r

1

505.2053 'f 500 1

— 1 ]—*Ί 1 I 1— 600

Figure 1. Scan showing unknown contaminant.

In order to determine the source of the contaminant, the components of the liquid chromatograph were systematically removed from the solvent flow path while the contaminant level was monitored. After seeing no change in contaminant signal after all of the LC components except the pump were removed, methanol was infused into the mass spectrometer using a syringe pump. As the signal was still observed, the L C pump was removed from consideration. This left only two possibilities for the contamination: the

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

99 methanol or the mass spectrometer. After infusing methanol from several different manufacturers and still observing the contaminant, it was concluded that the source of the contamination was the mass spectrometer. Due to the high level of contamination, identification of the contaminant was necessary in order to determine its source with the expectation that it could possibly be removed.

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

Methodology A l l analyses were performed on an LCMate reverse-geometry, doublefocusing mass spectrometer (JEOL USA, Peabody, MA). The mass spectrometer was equipped with an electrospray ionization (ESI) source and a linked scan collision chamber. The liquid chromatography system consisted of an Agilent G1312A binary pump and an Agilent G1327A autosampler (Palo Alto, CA). Manual injections were made using a Rheodyne 7725i injection valve (Rohnert Park, CA). Infusions were performed using a single-syringe infusion pump (Cole Parmer, Vernon Hills, IL). Octylphenol (OP) and nonylphenol (NP) were provided by Dr. Carter Naylor (Huntsman Corporation, Austin, TX). HPLC grade or higher methanol was purchased from Burdick and Jackson (Muskegon, MI), Fisher (Pittsburgh, PA), and E M Science (Gibbstown, NJ). A l l calculations were performed using LCMate 2000 Data Reduction Version 1.9v software (Shrader Analytical Laboratories, Detroit, MI). A l l analysés were performed using electrospray ionization in the negative mode.

Accurate Mass Measurement For any sector mass spectrometer, the following equation applies:

where m/z is the mass/charge ratio, k is a constant, Β is the magnetic field strength, and V is the accelerating voltage. This equation implies that either the magnetic field strength or the accelerating voltage may be scanned while the other is held constant in order to determine the mass of an ion that enters the mass spectrometer. Although both techniques can be used on a double-focusing mass spectrometer, an accelerating voltage scan is typically used for accurate mass measurements. This type of scan is linear and can therefore be calibrated with only two points. In addition, there is no hysteresis as with a magnetic field scan. In order to get the best mass resolution possible

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

100 during an accurate mass measurement, the appropriate slits are set to their smallest widths to select for the mass of interest. In order to calibrate an accurate mass measurement, an appropriate calibration mass set must be measured with the compound of interest. Since the unit resolution spectrum showed a mass for the contaminant of approximately 212, OP and NP were chosen due to their [M - H]" exact masses of 205.1592 and 219.1749, respectively. An injection of 25μΙ, of a lppm solution of OP and NP was made. A summary of the mass spectrometer conditions is shown in Table I. After several scans were obtained, the octylphenol and nonylphenol peaks were assigned their appropriate exact masses. Using this calibration, the mass of the unknown compound was determined to be 212.0734. A scan from the accurate mass measurement is shown in Figure 2.

Table I. Accurate Mass Measurement Conditions Parameter Orifice 1 Voltage Ring Lens Voltage Desolvation Plate Temperature Orifice 1 Temperature Main Slit Alpha Slit Collector Slit Multiplier Centroiding Method Scan Speed Magnetic Field Strength Nebulizer Gas Drying Gas

Value 10 V 40 V 250°C 80°C 5000 1.0 15μηι 700V Moments 3 sec/scan 2500 On Off

Elemental Composition Determination Once an accurate mass measurement has been carried out, the value obtained can be used to determine the elemental composition of the corresponding compound. Although the mass of the compound in this study is relatively small, the number of possible elemental compositions can be quite large. In order to narrow down the list of candidates, several parameters used in the calculation algorithm must be defined.

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

101 212EXA~1 Measure 212 exact mass Scan: 8 TIC=175988 Base=11.3%FS

#ions=180 RT=2:35

212.0734

100-

80-

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

60%

-

40-

20213.0757 214.0703 m/z208

—I— 210

212

214

216

Figure 2. Scanfromaccurate mass measurement of unknown compound.

The first parameter to be considered is the allowable error for the calculation. Typically, a mass error of 10 mmu or less is considered reliable using an instrument with a resolving power of 5000 full-width half-maximum. Consequently, 10 mmu was used for this calculation. The range of the unsaturation value for a compound is the next parameter to be considered. The unsaturation value is a way of expressing the number of rings and sites of unsaturation on the compound of interest. Since the smallest unsaturation value for an organic compound is -0.5, this was used as the lower limit. Unless a large number of double bonds or rings is expected, a value of 20 is safely reasonable for small organic molecules. This value was used as the upper limit. The final value to be considered is the number and type of elements allowed in the calculation. This will vary greatly depending on the particular application being considered. For the purposes of this application, the standard elements carbon, hydrogen, nitrogen and oxygen were allowed. In order to identify other possible elements, the spectrum from the accurate mass determination was closely examined. This spectrum can be seen in Figure 2. The characteristic chlorine-3 5/chlorine-37 and bromine-79/bromine-81 pairs were not observed, therefore they were eliminated from consideration. Since phosphorus is a relatively uncommon element of environmental contaminants, it was also removedfromconsideration.

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

102 Upon further review of the spectrum, an atypical [M - H + 2]" peak was observed. Typically, for larger molecules, the [M - H + 2]" would gradually become larger as the possibility of two carbon-13 atoms on the molecule increased. However, for a nominal mass of 213, the expected percentage of two carbon-13 atoms would be only about 0.1%. This is not in agreement with the observed [M - H + 2]' peak which is approximately 5% of the base peak. Since the most abundant isotope of sulfur, sulfur-34, has a natural abundance of 4.5%, it was considered a viable candidate for the [M - H + 2]" peak. Therefore sulfiir was included in the calculation. When the calculation was complete, a table showing the possible formulae, their unsaturation value, and their error was generated. This is shown in Table Π. In order to narrow down the possible formulae, several techniques may be used. For this application, the nitrogen rule and the even/odd electron rule were of the most value. The nitrogen rule states that a compound with an even nominal mass will have an even number of nitrogen atoms and a compound with an odd nominal mass will have an odd number of nitrogen atoms. Since the compound of interest was shown to have a nominal mass of 213, the nitrogen rule states that the compound must have an odd number of nitrogen atoms. When the nitrogen rule is applied to Table II, the list can be narrowed down to two formulae, C H NO and C H NO S. The even/odd electron rule applies to the formation of ions in a mass spectrometer. Ions can be formed with either an even or an odd number of electrons. During soft ionization processes, such as electrospray ionization, only even electron ions are formed. Thus, due to the nature of the unsaturation calculation, the unsaturation value for an ion formed using electrospray ionization will have a remainder of 0.5. When applied to Table II, the list can be narrowed down to two formulae, C H NO and C H NO S. These results match those discussed above using the nitrogen rule. 13

10

2

10

14

2

13

10

2

10

14

2

Table Π. Elemental Composition Results Formula C H N 0 S CH N0S C,H N 0 S C H N 0 C H NO S CnH 0 S C]iH, S C H N 0 C H NO C H 0 C H S 6

6

9

I0

16

2

16

2

4

12

2

2

12

2

14

16

2

6

2

12

8

10

I4

4

2

13

14

2

12

12

2

2

2

2

2

Unsaturation 0 0 5 5 4.5 4 4 10 9.5 9 9

Error (ppm) 8.1

-9.7 11

-6.3 -1.1 -14 4 15 2.2 -10 7.4

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

103 In order to eliminate one of the remaining two formulae, the accurate mass spectrum was again considered. Since the formula C H NO does not explain the unusual [M - H + 2]~ peak, it was discarded in favor of C H NO S which includes a sulfur atom. This leaves a formula of C H NO S for the uncharged, parent compound. 13

10

2

10

10

15

14

2

2

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

Linked Scan MS/MS Although the number of molecular formulae has now been narrowed down to one, the possibility of several isomers still exists. The structural possibilities of the most commonly used isomers are shown in Figure 3. In order to determine which of these structures is correct, linked scan MS/MS was performed on the compound. This was done because of the ability of linked scan MS/MS to provide structural information and an accurate fragment mass. High energy, linked scan MS/MS derives its name from the high energy to which the ions of interest are accelerated and the 'linking" of the two sectors of the mass spectrometer during scanning. When a precursor ion is accelerated to a kinetic energy of approximately 1 keV or higher, the collision that occurs with the target mass is considered high energy. At this energy, excited electronic states within the precursor ion are produced. These types of collisions produce a very broad internal energy distribution that leaves virtually all structurally possible fragments with some probability of occurring. Since the center-of-mass energy of the ion is such a small percentage of the much larger kinetic energy, the target mass does not have nearly as much effect on the MS/MS spectrum as would a low energy collision. This means that parameters such as collision energy, temperature, and pressure do not have a profound effect on the MS/MS spectrum, leading to more reproducible results.

A

Β

C

Figure 3. Structures ofpotential isomers. A) p-butylbenzenesulfonamide B) 4-dimethylaminobenzyl methyl sulfone C) N-butylbenzenesulfonamide

In Liquid Chromatography/Mass Spectrometry, MS/MS and Time of Flight MS; Ferrer, I., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2003.

104 During a linked scan of a double-focusing mass spectrometer, the magnetic and electric sectors are scanned simultaneously while being held at a constant ratio. Since the magnetic sector separates ions according to their momentum (mv) and the electric sector separates ions according to their kinetic energy (l/2mv ), the ratio of the two is inversely proportional to their velocity. This is shown in the formula below: 2

Downloaded by UNIV ILLINOIS URBANA on June 3, 2013 | http://pubs.acs.org Publication Date: May 20, 2003 | doi: 10.1021/bk-2003-0850.ch006

Ratio of Sectors = All ions leaving the ion source are accelerated to the same kinetic energy. Since the kinetic energy of the ions is l/2mv , ions with different masses will have different velocities. In the field-free region after the ion source, the accelerated ions are introduced into a collision cell filled with a collision gas, in this case helium. The precursor ions collide with the helium to produce fragment ions. Assuming the velocity of the ion does not change when it fragments, the velocity of the product ion will be the same as the precursor. As was shown above, performing a linked scan of the two sectors selects for velocity. Thus, linked scan MS/MS selects only product ions that have the same velocity as their precursor ions. This process was carried out for the unknown compound. A summary of the experimental conditions is shown in Table III. The resulting linked scan MS/MS spectrum for the compound of interest is shown in Figure 4. The measured mass of the fragment, 140.9968, corresponds very well (