Analytica Chimica Acta, 119 (1980) 129-135 0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
A DOUBLE QUADRUPOLE SPECTROMETRY
FOR MASS SPECTROMETRY/MASS
and R. G. COOKS*
of Chemistry. Purdue University. West Lafayette,
W. J. FIES Finnigan Coworation,
Sunnyvale. CA 94086
(Received 3rd March 1980)
SmlMARY A double quadrupole mass spectrometer has been constructed to study unimolecular and collision-induced dissociation products from mass-selected ions. The two quadrupoles are closely coupled and the dissociation products sampled from a 2.5-mm interquadrupole region. Spectra obtained on the double quadrupole instrument are compared with published data obtained with triple quadrupole and reversed-sector (MIKE) mass spectrometers. The results indicate that the simple double quadrupole spectrometer is a highly efficient device which is a viable alternative to more complex quadrupole or sector instruments for obtaining dissociation spectra of mass-selected ions.
Bi- and multi-analyzer mass spectrometers have been in use for many years [ 11. Perhaps the commonest configuration has been that in which an electric and magnetic sector are operated together to achieve high mass resolution. Uncoupling these analyzers so that they act independently and sequentially upon an ion beam is a more recent development. Trace organic analysis by the technique of sequential mass analysis has been accomplished over the past five years [Z] by using a reversed-sector mass spectrometer (BE, magnetic field followed by electric field). These studies employed a mass-analyzed ion kinetic energy (MIKE) spectrometer  to separate individual ionized components’ and to characterize their fragments generated by collisioninduced dissociation of the selected ions. The concept of ionization, ion separation, dissociation and fragment mass analysis as a method of mixture analysis has been adapted to quadrupole instruments in the form of a triple quadrupole (Q&Q)device [4,5]. In this instrument the center quadrupole is employed as a focusing device to contain the dissociation products formed in the low-energy (eV range) ion-molecule reactions. In this and subsequent papers, the use of a double quadrupole (QQ) instrument is explored for rnas spectrometry/mass spectrometry (m.s./m.s.) , i.e. for trace organic analysis without prior chromatographic separation and also for studies on the chemistry of individual ions selected from an ion mixture. Of particular interest is a comparison of the spectra obtained by
QQ with those generated by MIKE spectrometry on the one hand and by QQQ on the other. Both unimolecular and collision-induced dissociations are discused. The former occur at longer times than thee observed in typical sector instruments; the latter occur at about a thousand times lower axial energy than the analogous reactions in a sector instnunent. In spite of these differences, very similar spectra will be shown to result (for electron ionization). EXPERIMENTAL
A schematic diagram of the configuration of the QQ mass spectrometer is given in Fig. 1. The quadrupoles are standard Finnigan units, with rods (3-mm radius, 12.7-mm long) operated by using two Finnigan model 3000 electronic modules and a model 1015 preamplifier and display unit. The sowcc is a Finnigan electron impact source, configured for vepor introduction and for probe introduction of the sample into the vacuum system. The detector is a Cal&o Channeltron model 4700. The quadrupole 855emblies are mounted in a holder such that the rods are separated by a grounded plate which is normally 2.5mm thick with a centrally located aperture (6 mm diameter). In later experiments, target gas was introduced via a channel drilled radially through this aperture plate. The entire stmchlre, however, is very open and the sample introduced into the source makes a considerable contribution to the target gas pressure. Pressure was read on an externally mounted ion gauge and was typically 6 X lo-’ torr. Higher pressure would have given more collision-induced dissociation but this was
- . II -
. !. \ \
Schematic cell. CGI.
the double qurdrupok ~rutrument: 1s. ion source; gas Inlet; VS. to ncuum system. 0, detector.
precluded because of the lack of differential pumping and the danger of arcing in the multiplier. An m.s./m.s. spectrum is obtained by setting Q2 to pass all masses in the r-f.-only mode and then SCarming Q1 to locate the ion of interest. This ion is then transmitted at a fixed Q1 setting and Q2 is scanned in the mass analysis mode to obtain the m.s./_m.s. spectrum. R?ZXJL’Li’SAND
Studies on the diisopropyl
Figure 2 compares the dissociation spectrum of 114’,
the molecular ion of diisopropyl ketone, with the MIKE spectrum of the same ion. Both spectra were taken in the presence of target gas (the sample itself and nitrogen, respectively). The agreement is excellent, the only difference being in main beam intensity. The transition 114+ + 71+ is known [ 71 to be a collisioninduced dissociation in MIKE spectrometry while 114’ + 70’ is known to be unimolecuiar. These features also hold true in the QQ experiment, as the data of Fig. 3 indicate. This result demonstrates that substantial unimolecular de00
120 Pressure ( x IO+
Fig. 2. Comparison of the fragments generated at high pressure from the diisopropyl ketone molecular ion (114+) using the double quadrupole (upper) and the MIKE spec-
trometer (lower). Fig. 3. Linear plot for the relative yields of 71’ and 70+ (from 114+, diisopropyl ketone) against pressure. The individual reactions 114+ -+ 71+ and 114+ + 70+ are first andzero order. respectively, in pressure_
composition occurs some 30 ps (the transit time through Q #) after ionization. While it has not been directly demonstrated, one assumes that some of the dissociations being sampled in the QQ experiment occur within the quadrupoles, near the interquadrupole region. The diisopropyl ketone system was used to study the effects of instrumental and experimental variables on the m.s./m.s. spectrum. The diisopropyl ketone fragmentation processes did not change significantly except in relative main beam to fragment intensity as the interquadzupole spacing was increased from 1.6 mm to 8.4 mm, indicating poor fragment ion collection efficiencies at large spacings. The fragment ion collection efficiency and the overall transmission of the main beam ions were found to increase as the interquadrupole distance was decreased. In routine operation, Q, was operated at 2.14 MHz and Q, at 1.85 MHz. The frequency of Q, was varied from 2.49 MHz to 1.75 MHz; again there were no major changes in the relative intensities of the fragments in the spectrum. However, when Q, and Q2 were operated at nearly identical frequencies the fragment ion signals were observed to show a 5-10% intensity modulation. The modulation frequency increased as the frequency difference between Q, and Qz increased, and may be the result of r-f.-phase discrimination during ion entry into Q2. This modulation was not observable when Q, was operated at 2.14 MHz. Comparison of spectra obtained on double quadrupole and triple quadrupoie spectrometers Comparisons were also made between spectra obtained on the double quadrupole spectrometer and published data for the triple quadrupole [ 71. Some of these results are shown in Figs. 4-6. The QQ spectra for these figures were obtained without use of an added collision gas. The sample served as the collision gas and the instrument pressure was raised to the values indicated by increasing the sample gas flow rate. Except for the scaling of the main beam of ions, the agreement is excellent. Only two minor differences occur: (1) the QQQ spectrum of cyclohexanone molecular ion shows 81’ which is absent in the QQ spectrum; (2) the QQ spectrum of benzene shows 63’ which is much less intense in the QQQ spectrum. These differences probably reflect target gas and axial ion kinetic energy effects. For example, the cyclohexanone QQQ spectrum shows 99*, i.e. an (M + H)’ ion which is probably a consequence of the target gas used in that experiment. Protonated cyclic ketones should undergo extensive dehydration; this is the base peak of the MIKE spectrum for protonated cyclopentanone , and explains the presence of the 81’ ion in the QQQ spectrum. In later experiments, gas was introduced into a collision cell located between the quadrupoles. The pressure of the celI was monitored by using a capacitance manometer (Baratron 170-M). Collision cell pressures of 2-4 m torr were found to give about a five-fold increase in the absolute intensity of the fragment ion spectra when compared with spectra obtained without collision gas but at the same sample pressure. At a pressure of 3 mtorr the parent ion of diisopropyl ketone (m/z 114.) was attenuated to approxi-
I 70 m/z
Fig. 4. Comparisons of the dissociations of n-hexane molecular ion (86+) recorded double quadrupole (sample = target) and the triple quadrupole (Ar target ) [ 91. Fig. 5. Comparison of the m.s./m.s. spectra of cyclohexanone ded using double and triple quadrupole maSS spectrometers.
mately 15% of the intensity observed in the absence of collision gas (both Q, and Q2 in mass-analysis mode). Presumably most of this loss is due to scattering. Figure 7 shows the m.s./m.s. spectrum of diisopropyl ketone m/z 114+ using air as a collision gas where both the fragment ions and the parent ion were recorded on the same amplifier range setting.
Transmission efficiency In the present configuration the transmission efficiency of Q, is estimated to be 10-1570. The transmission efficiency of Qz is also estimated at lo15%. These values were determined by operating one quadrupole in the mass-analysis mode and the other in the r.f.-only mode. With the massanalyzing quadrupole set to pass a given mass, the detector current was measured. The r.f.-only q-uadrupole was then switched to the mass-analysis mode and the detector current measured again, now with both quadrupoles performing mass analysis. The detector current decreased to lO-15% of the current for an m.s./r.f.-only or r.f.-only/m.s. configuration when going to an m.s./m.s. configuration (without collision gas).
Fig. 6. Comparisoo of the m.r/m.s. spectra of C,H;‘rvcordcd quadrupolc mas5 specbometers. Fig. 7. M.s./m.s. spectra of diisopropyl (2.3 mtorr tir).
using the double aod triple
for a presurized
Transmission efficiency of ions from the exit of Q, to the entrance of Q2 is estimated at approximately 50%. This value was determined by comparing detector currents in experiments which involved closecoupling of the two quadrupoles without a collision cell with the detector currents observed using the present collision cell. The transmission effkiency from Q, to Q1 varies with the distance of separation, lensing, and/or aperture sizes of the particular collision cell used. The fragment ions which reached the detector in Fig. ‘7 represent 10% fragmentation of the 114’ ion in the collision region. Because of the lO15% transmission efficiency of Q1, approximately 1% of the 114’ ion current exiting Q, was collected as detectable fragment ion current. Expressed differently, some 20% of the ion current arising from 114’ recorded in the absence of collision gas and in the double mass-analyzing mode is collected end recorded in Fig. 7. This number compares favorably with that reported for the triple quadrupole (41. Continuing experiments are underway to optimize the collision-region configuration and improve the collection efficiency of ions in the presence
of collision gas. The resuIt.ssuggest that the simple double quadrupole instrument described here is a viable alternative to more complex quadrupole or sector instruments in obtaining dissociation spectra of selected ions. These capabilities are explored further in subsequent papers. This work was supported by the National Science Foundation CHE 7701295.
REFERENCES 1 J. H. Futrell and T. 0. Tiemsn, in J. L. Franklin (Ed.). Ion Molecule Reactions, Plenum, New York, 1972, Ch. 11; E. Lindhoirn, in J. L. Franklin (Ed.), Ion Molecule Reactions, Plenum, New York, 1972, Ch. 10: J. P- L’Hote, J- Cb. Abbe, J- M. Paulus and R. Ingersheim, Int J. Mass Spectrom. Ion Phys., 7 (1971) 309; M. L. Vestal and J. H. Futreii, Chem Phys. Lett., 28 (1974) 559. 2 T. L. Kruger, J_ F. Litton, R. W. Kondrat and R. G. Cooks, Anal. Chem., 48 (1976) 2113;R. W. Kondrat and R. G. Cooks, Anal. Chem., 50 (1978) 81A. 3 J. H. Beynon, R. G. Cooks, J. W. Amy, W. E. Bsitinger and T. Y. Ridley, Anal. Chem., 45 (1973) 1023A. 4 R. A. Yost and C. G. Enke, J. Am. Chem. Sot., 100 (1978) 2274; Anal. Chem., 51 (1979) 1251A. 5 D. F. Hunt and J. Sbabanowitz, 27th Ancual Conference on Mass Spectrometry and Allied Topics, Seattle, Washington, June 1979, paper FAMOCQ. 6 F. W. McLafferty and F. M. Bockhoff, Anal. Chem., 50 (1978) 69. 7 J. F. Litton, T. L. Kruger and R. G. Cooks, J. Am. Chem. Sot.. 98 (1976) 2012. 8 M. L. Sigsby, R. J. Day and R. G. Cooks, Org. Mass Spectrom., 14 (1979) 273. 9 R. A_ Yost, C. G. Enke, D. C. McGilvery. D. Smith and J- D. Morrison. Int. J. Mass Spectrom. Ion Phys., 30 (1979) 127.