molecule reaction studies in the center quadrupole of a triple quadrupole mass spectrometer

International Journal of Mass Spectrornetry and Zon Processes, 82 (1988) l-15 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

1

ION-TRAPPING TECHNIQUE FOR ION/MOLECULE REACTION STUDIES IN THE CENTER QUADRUPOLE OF A TRIPLE QUADRUPOLE MASS SPECTROMETER

G.G. DOLNIKOWSKI,

M.J. KRISTO,

C.G. ENKE

and J.T. WATSON

*

Department of Chemistry, Michigan State University, East Lansing MI 48823 (U.S.A.) (First received

21 March 1987; in final form 7 August

1987)

ABSTRACT A novel ion-trapping technique has been developed to facilitate the detection of reaction products formed in the central quadrupole of a triple quadrupole mass spectrometer. Ions that are produced by ion/molecule reactions are trapped for a variable period of time in the central quadrupole and are then pulsed toward the detector by means of an extraction lens. Full MS/MS scans using the repetitive trap and pulse technique have been achieved with no changes to the instrument hardware and only minor software changes in a FORTH-based microprocessor data system for the instrument.

INTRODUCTION

The triple quadrupole mass spectrometer (TQMS) was developed initially to study laser photodissociation reactions in the region of the center quadrupole [1,2]. Researchers soon discovered that, by using the central quadrupole as a collision chamber, it was also possible to study collision-induced dissociation (CID) [3] and ion/molecule reactions [4,5]. Of these three reaction modes, CID using nitrogen or argon has been used most frequently in scientific applications of the TQMS. Very recently, however, there has been renewed interest in using the TQMS to study ion/molecule reactions in the central quadrupole region [6-121 since interactions of ions with a reactive collision gas may provide different information about analyte ions than interactions with an inert gas [7]. Unfortunately, the majority of ion/molecule reactions studied in the center quadrupole of the TQMS appear to produce low yields of product ions, a phenomenon probably related to the lack of kinetic energy possessed by these thermalized product

* To whom correspondence

0168-1176/88/$03.50

should be addressed.

@ 1988 Elsevier

Science Publishers

B.V.

2

ions. Therefore, products of ion/molecule reactions in the central quadrupole are often not detected because they tend to remain in the central quadrupole region. In addition to this difficulty, the standard TQMS provides poor control over the residence time of ions inside the central quadrupole 1121.The standard technique to alleviate this difficulty is to use an ion extraction lens at the exit of the central quadrupole. Ion currents for thermalized ions in the central quadrupole, however, are limited by the rate of diffusion along the axis of the quadrupole from the point where their axial kinetic energy becomes zero until they reach the drawout field which extends only a short distance into the inter-rod space from the exit end. This paper describes a novel approach for varying the average residence time of ions inside the central quadrupole and for improving the intensity of signals representing ion/molecule reaction product ions. This method stores product ions in the central quadrupole, then extracts those ions present in the exit region of the central quadrupole. An instantaneous current many times greater than the steady-state current is obtained. This technique requires no hardware modifications to the instrument and can be applied to most TQMS instruments currently available. The technique of ion trapping in a TQMS is illustrated in this paper with a reaction of methyl cations with neutral acetone. This reaction produces ten primary products as characterized by ICR spectrometry [13]. The particular ion/molecule process used to characterize the effects of ion trapping in the central quadrupole is the reaction of protonated glycerol with neutral acetic acid. Similar ion/molecule reactions between protonated alcohols and acetic acid have been investigated by ICR spectrometry [14]. The reaction used to demonstrate the increase in signal-to-background noise ratio enhancement is that of the protonated molecule of acetone with acetone. EXPERIMENTAL

Instrzdmentation The instrument used for these studies is a prototype triple quadrupole mass spectrometer built by Extrel Inc. (previously Extranuclear Labs) 1151. The instrument has a dual EI/CI ion source that has been modified to permit fast atom bombardment (FAB) ionization. The fast atom gun is a capillaritron probe gun from Phrasor Scientific [16]. All of the samples used in this study were introduced either through a volatile liquids inlet or by a direct probe. Collision gases were admitted into the collision region of the TQMS by means of a stainless steel vacuum line. The pressure of gas in the collision region was regulated by a model 216 pressure/flow controller from Granville-Philips, Inc. [17].

TABLE 1 Typical

conditionsfor ion trapping in the center quadrupde of the TQMS

Parameters

EI

+c1

-CI

FilWlMM Repeller CI El Ll L2

70 eV 14.3 v 2.7 v 14.9 v 23.7 V - 30.0 v

100 eV 33.0 v 32.2 V - 185.8 v - 3.7 v - 13.0 v

100 - 46.6 - 37.6 - 1.5 75.9 - 23.3

FAB eV V V v v V

0.0 0.0 - 106.3 - 26.4 18.9

v v v V V

;

-0.5 10.0 v

- 191.3 20.0 v

- 110.4 33.0 v

v - -9.4 72.2 V

:

14.0 5.9 v

V 32.2 4.0 v

- 66.9 26.6 v V

V -1.8 5.0 v

100.0 4.0 v -1.7 kV - 3.0 kV

100.0 20.0 v - 1.7 kV - 3.0 kV

- - 100.0 32.7 v V -2.0 kV 3.0 kV

: Multiplier Conversion Dynode FAB Gun

- 100.0 13.2 - 1.7 - 3.0 10.0

V v kV kV kV

The ion path consists of a standard EI/CI source, a three-element einzel lens (Ll, L2, L3), quadrupole 1 (Ql), an interquad aperture lens (L4), quadrupole 2 (Q2), an interquad aperture lens (L5), quadrupole 3 (Q3), and an off-axis continuous dynode electron multiplier with a conversion dynode. In the ion-trapping mode, L5 (used as the trap control lens) was maintained at + 100 V for positive ions and - 100 V for negative ions. After trapping, the opposite potential was applied to L5 to produce the extraction pulse. Typical TQMS instrument parameters for EI, CI, and FAB are shown in Table 1. An 8086-based microcomputer controls the TQMS and acquires data using a software control system built on the programming language FORTH. The control system, based on the work of Myerholtz [18] and adapted by Extrel, Inc., provides modular software which allows considerable flexibility in using the instrument. In addition, the system can be programmed by the user. The authors easily modified the basic data collection algorithm to permit the trap and pulse data collection technique described in this paper to be used with daughter scans, parent scans, and neutral loss/gain scans. Chemicals Glacial acetic acid was obtained from Mallinckrodt [19] as an analytical reagent; acetone was obtained from Fisher Scientific [20]. Both were used without further purification. The glycerol used in this project was vacuumdistilled. Xenon obtained from Matheson 1211 was 99.99% pure.

Ion /molecule

reaction conditions

The reaction of the methyl cation and acetone was carried out in the TQMS by introducing 5 x 10m6 torr of acetone into the ion source and 5 X lop4 torr of acetone into the central quadrupole region. The acetone in the ion source region was ionized by 70 eV EI, which produced the methyl cation in addition to many other fragment ions. The methyl cation was mass-selected using the first aquadrupole and reacted with the neutral acetone in the central quadrupole. The reaction of protonated acetone and acetone to form the proton-bound dimer was carried out by ionizing acetone (5 X 10W4 torr) in the CI source with 300 eV electrons. The protonated acetone was mass-selected by the first quadrupole and reacted with acetone (5 X low4 torr) in the second quadrupole. The reaction of protonated glycerol and acetic acid was implemented by protonating the glycerol during FAB in the ion source of the TQMS and by regulating the partial pressure of acetic acid in the center quadrupole region. The ion source and central quadrupole regions are differentially pumped so that there is little mixing of the neutral vapors of glycerol and acetic acid. The protonated glycerol was mass-selected by the first quadrupole and reacted with the acetic acid in the second quadrupole to form the protonbound adduct ion which, in turn, was mass-selected by the third quadrupole. RESULTS AND DISCUSSION

Ion trapping in the central quadrupole

All experiments were conducted with a standard configuration triple quadrupole mass spectrometer [22]. During a continuing study of ion/molecule reactions in the center quadrupole of the TQMS, it was .observed that the signal intensity for ion/molecule product ions increased dramatically for several milliseconds when the potential on the ion trap control lens (L5, the interquadrupole lens between the second and third quadrupoles) was decreased abruptly. For positive ions, the increase in peak intensity was greatest if the potential on L5 was increased to a value at least 10 V more positive than the d.c. offset voltage of the central quadrupole. Previous workers have used a large steady voltage on this lens to extract ions that had lost their forward momentum from the central quadrupole. In these applications, L5 was sometimes called the extraction lens. In this work, where we use L5 to both trap and extract ions, we will refer to it as the trap control lens.

5

Apparently, the central quadrupole and its trap control lens (L5) can act as an ion trap for ion/molecule reactant and product ions with low kinetic energies. A fraction of these trapped ions can then be drawn out toward the detector by means of a voltage pulse on L5. The maximum signal intensity of the resulting ion pulse may be many times greater than if ion/molecule product ions are extracted by L5 under the normal steady-state operating conditions of the TQMS. We call this technique of detecting ion/molecule product ions the trap and pulse method. Note that, in the trap and pulse method, parent ions are constantly being created in the ion source and transmitted to the collision chamber by the first quadrupole. Because the third quadrupole cannot be scanned over its full range during the millisecond duration of the ion pulse generated by the ion extraction lens, the trap and pulse method is necessarily limited to collecting ions of a single nominal mass-to-charge value per pulse. In a related technique, ions can be injected into the collision chamber as a discrete packet by gating the ion source; if L5 is at + 100 V, positive ions are trapped in the central quadrupole until they are drawn out by a - 100 V negative pulse on L5. This technique we call the inject, trap, and pulse method. The major difference between the two experiments is that trap and pulse admits parent ions continuously to the collision chamber, whereas inject, trap, and pulse admits only a discrete number of ions into the collision chamber. In both modes, the trap control lens (L5) can block ions with low axial kinetic energies from reaching the detector. The inject, trap, and pulse mode was used to determine the length of time ions can be trapped inside the central quadrupole. The organic ion/molecule reaction chosen for this study was that of methyl cations reacting with acetone (investigated previously by ICR [13]) to form the product ions listed in Table 2. It is possible to trap stable product ions such as those represented by peaks at m/z 59 (acetone + H)+ and 117 (acetone dimer + H)+ in the central quadrupole region for up to 20 s. The number of trapped ions drops off exponentially with trapping time. The time constant for this exponential decay (measured using the m/z 117 peak at 5 X 10e4 torr acetone) was 2.8 s. Attempts to perform the inject, trap, and pulse experiment on some of the other ion/molecule reaction products in Table 2 were unsuccessful (no ion currents detected above the background present in the instrument) until the acetone pressure in the central quadrupole was lowered to 7 X lop6 torr. At this lower pressure of acetone, the less stable product ions such as those of m/z 43 were not completely consumed by reactions with neutral acetone and, thus, could be trapped, albeit for short periods of time. The ion of m/z 43 was trapped for 100 ms before its signal faded into the background. If the trapped ion abundance is plotted versus the axial energy of parent

6

TABLE 2 Ion/molecule

products of the reaction between &hemethyl cation and acetone

m/z

Formula

15.0 29,O 31.0 41.0 43.0 45.0 57.0 58.0 59.0 117.0

CH; a C& b CH&+ b C,H; b CH$O+ b CH,OCH; b C3H50+ b C3H60+ b CJH,O+ b C,H,30; ’

a Parent ion. b Primary reaction product. ’ Product formed by reaction of primary products with acetone.

ions entering the central quadrupole, a sharp peak is obtained at approximately 0 eV. This is the same type of energy effect found for most ion/molecule reactions observed without trapping in the central quadrupole In the trap and pulse mode, parent ions enter the central quadrupole at a constant rate from the ion source. Therefore, a plot of the area under the ion pulse profile vs. trapping time will reveal trapping efficiency from its linearity; losses dependent on the ion concentration would cause a negative deviation from linearity in this plot. Regression analysis of the area under the pulsed signal profile of the protonated dimer of acetone of mass 117 (from the reaction of protonated acetone and acetone) versus trapping time indicated a trapping efficiency of approximately 99%. This indicates that this ion is stable over the 100 ms experiment period and that the central quadrupole of the TQMS can effectively trap almost 100% of the m/z 117 ions that are produced within it. The extraction efficiency is not lOO%, however. Given that the trapping efficiency is virtually loo%, the extraction efficiency can be calculated from the number of ions (amount of charge) contained in a given pulse (see Fig. 1) and the amount that would be expected given 100% trapping efficiency and 100% extraction efficiency. The amount expected from 100% extraction of the trapped ions is the steady-state ion current multiplied by the trapping time, assuming the ion in question is not a product ion or an unstable parent ion. As described in the section on data acquisition, exact determination of the integral charge present in a given pulse is difficult due to the low signal to background noise in the long tail of the pulse profile. However, the ion

7

current profile was integrated between visually established limits which were from 0 to 60 ms of the pulse. This pulse area (charge) value is artificially high by a small amount which corresponds to the unavoidable integration of the steady-state ion current during this period. The pulse area was then measured as a function of trapping time. A plot of the measured charge, Q, versus the trapping time, t, for each pulse yields a straight line with a slope dQ/dt and an intercept which is equal to the integrated steady-state current. The slope, dQ/dt, is the rate of accumulation of extractable and detectable charge in 42 during the trapping time. If the extraction efficiency were lOO%, dQ/dt would be equal to the steady-state ion current. The extraction efficiency was experimentally determined for benzene molecular ion, produced by EI in the source and thermalized in QZ with acetone collision gas. (Benzene molecular ion had previously been shown to be non-reactive with acetone under these conditions.) The extractable accumulation of benzene molecular ions was 19% of the integrated steady-state ion current under the same conditions. The trapping algorithms and their effects on data acquisition Figure 1 illustrates the time profiles of voltage on the third source lens (L3), the trap control lens (L5) voltage, and product ion current in the inject, trap, and pulse mode. In this mode, the voltage on L3 is used as a gate to control parent ions leaving the ion source region and entering the collision chamber. The voltage on the trap control lens (L5) is used to trap ions in the collision chamber and to pulse the stored ions toward the detector. Raising the L3 voltage does not store ions in the ion source of the TQMS; rather, it seems that parent ions are simply deflected from their original flight path. The profile of the ion current in Fig. 1 is characterized by an abrupt rise to a maximum in less than 1 ms followed by an exponential decay with a time constant of 4 ms. The time constant for changes in the trap control lens (L5) voltage power supply is also 4 ms. The profiles of ion current in Figs. 1 and 2 are constructed from data points that are collected as frequently as the data system can acquire them (2500 points per second), so as to be certain of detecting the pulse maximum. Some of the characteristics of the TQMS system in the trap and pulse mode are shown in Fig. 2 which compares a product ion current profile with a corresponding profile of trap control lens (L5) voltage. During the trapping period, no product ion current arrives at the detector, but following the drop in trap control lens (L5) voltage, there is an abrupt rise in the product ion current followed by a decay to the level of product ion current produced during steady-state conditions (when the trap control lens (L5) is left at - 100 V indefinitely]. Unlike the decay observed in the inject, trap,

e-INJECT--TRAP--PULSE-

ms TIME

Fig. 1. Plots of ion abundance versus time and L3 and L5 potentials versus time showing one full inject, trap, and pulse cycle.

and pulse mode, the decay of ion current in the trap and pulse mode is not cleanly exponential. The time constant of the decay increases with collision gas pressure (i.e. the ion current profile becomes wider during the pulse phase), which indicates that the shape of the ion current profile is controlled at least partially by diffusional effects. The high voltages shown in Figs. 1 and 2 for the trap control lens (L5) are not necessary to observe the effects described. A pulse of + 10 V to - 10 V (for positive ions) is often just as effective, but the higher voltages ensure that the maximum number of ions is trapped in the central quadrupole and that the exit region of the central quadrupole is purged of ions between storage intervals. Also noteworthy in Figs. 1 and 2 is the fact that the increase in ion current begins as the L5 voltage drops to approximately 0 V, which is very near the offset voltage of the central quadrupole. Therefore, as the maximum ions current passes through the third quadrupole, L5 is not yet set at an extreme voltage to impart excess kinetic energy to the ions and, as a result, does not degrade mass spectral peak shape (see also Fig. 3). It is apparent from Figs. 1 and 2 that the fall time of the pulsed ion current

9

-TRAP

PULSE-

--

100%

100.0

0.0

-100.0

E +Lr 0

130

Id0

ms

TIME

Fig. 2. Plots of ion abundance versus time and L5 potential versus time showing one full trap and pulse cycle.

profile is about 20 ms; therefore, detection of the peak (pulse) maximum will always be achieved within ca. 20 ms of pulse initiation. Using the FORTH-based software control system of the TQMS, conventional MS/MS scans were easily modified to allow successive ion-trapping and pulsed data collection at each mass value. Thus, a conventional MS/MS daughter scan becomes a “reaction product” scan. In reaction product scans, the ion source, detector, lens, and quadrupole potentials are set and the microcomputer then generates the sequence of voltage pulses to the lenses as illustrated in Figs. 1 and 2. At each mass increment, the magnitude of the maximum ion current is determined; a series of these trap and pulse cycles at consecutive increments (0.1 mass unit) along the mass axis establishes the mass spectral peak profile. That is, immediately after the drawout pulse on L5, the microcomputer sequentially acquires a specified number of data points to establish the ion pulse profile at a given m/z increment. The data point with maximum magnitude during data collection for that cycle is used as the ion current value for that mass increment in defining the profile of the mass spectral peak. Finally, a peak-finding algorithm processes the sequential values of ion current along the mass axis to determine the peak intensity/mass value pairs for each mass spectral peak.

10

Alternatively, the microcomputer can integrate the area under the ion current profile. The integration algorithm, however, has been determined to decrease the precision of the measurement relative to that of simply finding the maximum ion current during the pulse. This means that the peak height is more reproducible than the peak area. The reason for this is that the data are in the form of an asymmetric peak with a Iong tail. The tail approaches a baseline which represents the steady-state product ion current. The low value of the steady-state current is sensitive to minor fluctuations in collision gas pressure and to random noise, such as that produced by the FAR gun: Therefore, the basetine contains considerable noise and also varies from pulse to pulse, causing the peak area to be iess reproducible than the peak height. Reaction scans are necessarily slower than the corresponding conventional MS/MS scans because a finite time is required at each mass increment to accumulate ions in the trap and then to collect ion current profile data following the extraction p&e. Typical durations for reaction scans (50-500 u) vary from 1 to 2 min, depending on the reaction being monitored. Therefore, cohection of trap and pulse data over the complete mass range (typicahy m/z 50-500) cannot be achieved on the gas chromatographic time scale. One can coIIect limited trap and pulse data on the gas chromatographic time scale by using an MS/MS technique calIed selected reaction monitoring in which the first and third quadrupoles are set to pass ions of one specific mass-to-charge ratio so that the TQMS selects for only one particuIar parent ion and one particular product ion at a time. One trap and pulse cycle can be accomplished every 50-100 ms, Therefore, one can expect to obtain at least IO data points across even the most narrow capillary gas chromatographie peaks by means of seIected reaction monitoring in this mode.

In the trap and pulse mode, parent ions are accumuIated continuously and all products are trapped in the central quadrnpole. Therefore, ion-trapping time in the centraE quadruple is an important parameter reIated to the magnitude of the signal for ion/moIecule reaction produet ions. The peak intensity of the ion/molecule reaction product ion increases rapidIy with trapping time and then Ievels off. For stable product ions, this phenomenon reflects a limit to the accumuIation of ions in the centraI quadrupole region due to space-charge. The same effect is observed frequently in other iontrapping techniques [23]. SeIecting a parent ion with a significant abundance [such as A4 + H+ of gIycero1 (m/z 93) generated by FAR] causes a high rate

11

Fig. 3. Three mass-sweeps of Q3 over the isotope peaks of the proton bound dimer of acetcme (a) by conventional data collection (250 u s-‘); (b) by conventional data coilection with real-time signal averaging (5 u s-*1; and (cj by trapamid p&e data collection (5 u s-l). Sign&to-background noise (S/BN) was c&&ted as the peak height divided by the standard deviaticm of the baseline.

of parent ion influx into the centralquadrupoIe.Enthis case, thespace-charge limit can be reached in less than 100 ms of trapping time. The major advantages of the trap and pulse technique over the conventional scanning techniques are ihustrated in Figs. 3 and 4. Figure 3 shows three MS/MS sweeps over the peaks representingthe isotopic muhiphcity of the proton-bound dimer of acetone that was produced by the ion/molecule reaction of proton&xl acetone with acetone In the central quadrupole region. Sweep (a) was colhxted in the conventional mode by stepping the third quadrupole in 0.1 u intervals over the product ion peaks of interest with no signd averaging (scan rate of 250 u s-r). Sweep (b) was also cohected in the conventional mode but with 20 ms of re&thne signal averagingper data point (scan rate = 5 u s-l). Sweep (c) was cohected using the trap and pt.&e mode with 20 ms of ion-trapping time per trap and pulse cycle (scan rate = 5 u s-*). Signal-to-background noise (S/BN) values in Fig. 3 were calculated as the peak height divided by the standard deviation of the background on either side of the peak. Figure 3 demonstratesthat, at equivalent overah scan speeds, the trap and pulse technique can produce sign&to-background noise ratios that are at least three times higher than those acquired by conventional data acquisition using reaMme signal averaging over equivalent time intervrds.The higher sigmd-to-backgrmnd noise ratio idicates a Imer due for the minimum detectable amuunt; it does not necess&ly indicate a higher reprodueibihtyin the peak value obtained. Signal averaging in the conventional mode improves the sign&to-noise ratio by discriminatingagainstrandom noise while favoring the reproducibEe sign& (the standard deviation for the peak height and the background both improve with the square root of the number of averages),.The trap and pulse techniquedoes not averageout n&e in this way; it improves signal-to-back-

I 00% 93

=

17000

CONVENTIONAL

GH+ J

(A+GH)+ \

153

4

I I.

00

120

100

140

M/Z

.--

100%

._

TRAP

AND

=

6800C 153

PULSE

::

(A+GH)+

5 P s

(A+GH

:

(OH- ‘to)+ 75

L

,+

117

GW

: ‘; 0 -i a

-2H20

‘;A+c”

I _AcEnL)+ k0

J

(A+GH-H~~)+

II9 12,

Y

I

O%-

80

100

A

=

ACETIC

G

=

GLYCEROL

M/Z

120

140

ACID

Fig. 4. Two product ion scans of the ion/molecule reaction between protonated glycerol and acetic acid (parent ion is (M + H)+ of glycerol at m/z 93) by (a) conventional data collection and by (b) trap and pulse data collection. A = acetic acid; G = glycerol.

ground noise by increasing the magnitude of the signal. This it achieves by integrating ion current inside the second quadrupole as well as by increasing the reaction yield of stable product ions. If there was 100% trapping efficiency and no change in reaction yield, after the first few ms, S/BN would theoretically increase linearly with trapping time. Only peaks representing ions which are stable for the entire period of ion trapping will exhibit the full theoretical increase in S/BN. Also, the precision in the peak measurement under the trap and pulse mode can be improved only by averaging consecutive pulse heights or areas. Figure 3 also shows that the shape of the mass spectral peaks representing the ion/molecule products and the mass resolving power of the third quadrupole are not degraded by the trap and pulse technique.

13

Figure 4(a) shows a conventional third quadrupole MS/MS scan of the ion/molecule reaction products from protonated glycerol and acetic acid in the central quadrupole. Figure 4(b) shows a reaction scan of the same ion/molecule reaction using the trap and pulse mode of operation. Figure 4 illustrates that the appearance of the ion/molecule product spectrum can be significantly different depending on the scan technique employed. In general, the ratio of product ion peak intensities to the parent ion peak intensity is greatly enhanced with the trap and pulse technique, as one would expect, since parent ions are being converted to product ions continuously during the trapping period and therefore, accumulating to readily discernible levels. Some ion/molecule product ions show more enhancement than others because, at long trapping times, unstable product ions can decompose or react with the collision gas to form secondary ion products. Therefore, trap and pulse tends to enhance dramatically the signals of stable ions, but may actually decrease the relative intensity of signals from less stable ions depending on the conditions in the collision chamber. Figure 4 also shows that at 7 x 10V5 torr of acetic acid in the collision chamber, the trap and pulse technique allows ion/molecule reaction products to be observed that otherwise might not be detectable at that pressure. The peaks at m/z 110, 117, 129, 121, and 135 are not present, even at low intensity, in the conventional spectrum unless the pressure of acetic acid is increased by at least an order of magnitude. However, these ions can be observed in the FAB spectrum of a mixture of glycerol and acetic acid; similar products have been observed in a reaction of protonated alcohols and acetic acid using ion cyclotron resonance [14]. The ions represented by peaks at m/z 119 and 121 are the result of proton transfer to the acetic acid dimer; the peaks at m/z 117 and 135 represent new ions that are the result of neutral losses from the proton-bound adduct of glycerol and acetic acid. The gas-phase ion/molecule chemistry of alcohols and acids is currently under more comprehensive investigation in this laboratory. CONCLUSION

It is possible to trap stable ions with very low axial kinetic energies (such as ion/molecule reaction product ions) with up to 100% efficiency in the central quadrupole region of a triple quadrupole mass spectrometer (TQMS) with no physical modifications to the instrument. The capacity to trap ions efficiently in the central quadrupole has several significant advantages for studying ion/molecule reactions in the TQMS: lower pressures of reactive collision gas can be used, the overall signal level and the detectability of ion/molecule reaction products can be improved, and ion/molecule reaction products which are not easily observed by conventional TQMS data

14

acquisition can be detected. Pulsing the trap control lens, as described in this paper, can extract up to 19% of the trapped ions. The ion-trapping technique described here could be used for enhancing the ion signals from other types of reaction in the collision chamber of the TQMS, such as charge transfer or photodissociation. Conceivably, ion trapping could be employed in other multi-quadrupole instruments such as the BEQQ 1241 and the pentaquadrupole instrument 1251 developed recently for ion/molecule reaction studies. ACKNOWLEDGEMENTS ‘Iplis work was conducted with instrumentation provided in part by a contract with the Office of Naval Research (NOOO14-81-K-0834) and funds provided by the Agricultural Experiment Station and various colleges at Michigan State University, and supported by a grant from the Biotechnology Research Program of the Division of Research Resources of the National Institutes of Health (No. RR-00480). We thank Victoria McPharlin and Jo Dutson for assistance in preparing the manuscript. We also thank Professor John Allison for helpful discussions and Timothy Heath for assistance in instrumentation maintenance.

REFERENCES 1 DC. McGilvery and J.D. Morrison Int. J. Mass Spectrom. ion Phys., 28 (1978) 81. 2 M.L. Vestal and J.H. Futrell, Chem. Phys. Lett., 28 (1974) 559. 3 R.A. Yost, C.E. Enke, D.C. McGilvery and J.D. Morrison, Int. J. Mass Spectrom. Ion Phys., 30 (1979) 127. 4 J.D. Morrison, K. Stanney and J.M. Tedder, J. Chem. Sot. Perkin Trans. 2, (1981) 838. 5 J.D. Morrison, K. Stanney and J.M. Tedder, J. Chem. Sot. Perkin Trans. 2, (1981) 967. Michigan State University, 1982. 6 J.A. Chakel, Ph. D. Dissertation, 7 D.D. Fetterolf, R.A. Yost and J.R. Eyler, Org. Mass Spectrom., 19 (1984) 104. 8 R.A. Yost and D.D. Fetterolf, Mass Spectrom. Rev., 2, (1983) 1. 9 D.D. Fetterolf and R.A. Yost, Int. J. Mass Spectrom. Ion Processes, 62 (1984) 33. 10 J.P. Schmit, P.H. Dawson and N. Beaulieu, Org. Mass Spectrom., 20 (1985) 269. 11 J. Jalonen, J. Chem. Sot. Chem. Commun., (1985) 872. 12 J.P. Schmit, S. Beaudet and A. Brisson, Org. Mass Spectrom., 21 (1986) 493. D.A. Herold, T.A. Elwood and J.H. Futrell, J. Am. Chem. Sot., 99 (1977) 13 R.D. Smiti 6042. 14 P.W. Tiedemann and J.M. Riveros, J. Am. Chem Sot., 96 (1974) 185. 15 Extrel, Inc., Pittsburgh, PA 15238, U.S.A. 16 Phrasor Scientific, Inc., Duarte, CA 91010, U.S.A. Inc., Boulder, CO 80303, U.S.A. 17 Granville-Phillips, Michigan State University, 1983. 18 C.A. Myerholtz, PhD. Dissertation, St. Louis, MO 63134, U.S.A. 19 Malliickrodt, 20 Fisher Scientific, Detroit, MI 48151, U.S.A.

15 21 Mameson Gas Products, &caucus, NJ 07094, U.S.A. 22 R.A. Yost and C.G. Enke, in F.W. McLafferty (Ed.), Tandem Mass Spectrometry, Wiley, New York, 1983, pp. 175-195. 23 R.T. McGiver, Jr. and R.L. Hunter, Int. J. Mass Spectrom. Jon Processes, 64 (1985) 67. 24 J.N. Louris, L.G. Wright, R.G. Cooks and A.E. Schoen, Anal. Chem., 57 (1985) 2918. 25 C. Beaugrand, G. Devant, D. Jaouen, N. Morin and C. Rolando, Presented at the 34th Annual Conference on Mass Spectrometry and Allied Topics, Denver, CO, May 1987, Paper MPB 34.