9
International Journal of Mass Spectrometry and Ion Processes, 108 (1991) 65-73 Elsevier Science Publishers B.V., Amsterdam 65 Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential S.M. Sadat Kiai, Y. Zerega, G. Brincourt, R. Catella and J. Andre Physique des Interactions Ioniques et Mol&ulaires (U&t! de Recherche AssociPe au C.N.R.S. No. 773), Laboratoire des Interactions Ioniques, VniversitP de Provence. Centre de Saint JPrBme--Case A61, F13397 Marseille Ckdex 13 (France) (First received 21 March 1991; in final form 17 April 1991) ABSTRACT In continuation of a previous theoretical work, we confirm experimentally the possibility of confining ions in a r.f. quadrupole trap supplied with an impulsional voltage. We have obtained the stability diagram of the trap corresponding to confined Xe+ ions. This diagram shows a zone of optimum confinement. Then, we have compared the confinement efficiency of the trap, when supplied by a sinusoidal or impulsional voltage, for two equivalent points of trap operation located in their corresponding stability diagram. INTRODUCTION In a previous paper [I], we have shown the possibility of confining ions inside a r.f. quadrupole trap supplied with an impulsional voltage of the form v, cos (Qt) 1 - k cos (2Qt) (1) where R/27c is the frequency of the r.f. field, and with 0 d k < 1. The advantage of such a signal, compared with the classic sinusoidal signal of trap operation, is that it presents periodic zero potential temporal zones during which it is possible to inject ions or electrons inside the trap for collisional studies without changing their initial energy before collisions. In addition, this signal has a narrower frequency spectrum than signals made with suitable square temporal components. So, it can be electronically realised without distortion and remains analytically known during experiment. We give here the first experimental results of confinement of Xe+ ions in a trap supplied with the voltage defined in eqn. 1. The stability diagram of the trap has been determined and we have compared the efficiency of confinement for equivalent points characterized by the same values of fl,,in the two stability diagrams corresponding to sinusoidal and impulsional excitations of the trap. This comparison was made for fi, = 0.5 (best confinement efficiency) 0168-1176/91/$03.50 0 1991 Elsevier Science Publishers B.V.

Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

Embed Size (px)

Citation preview

Page 1: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

International Journal of Mass Spectrometry and Ion Processes, 108 (1991) 65-73

Elsevier Science Publishers B.V., Amsterdam

65

Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

S.M. Sadat Kiai, Y. Zerega, G. Brincourt, R. Catella and J. Andre Physique des Interactions Ioniques et Mol&ulaires (U&t! de Recherche AssociPe au C.N.R.S. No. 773), Laboratoire des Interactions Ioniques, VniversitP de Provence. Centre de Saint JPrBme--Case A61, F13397 Marseille Ckdex 13 (France)

(First received 21 March 1991; in final form 17 April 1991)

ABSTRACT

In continuation of a previous theoretical work, we confirm experimentally the possibility of confining ions in a r.f. quadrupole trap supplied with an impulsional voltage. We have obtained the stability diagram of the trap corresponding to confined Xe+ ions. This diagram shows a zone of optimum confinement. Then, we have compared the confinement efficiency of the trap, when supplied by a sinusoidal or impulsional voltage, for two equivalent points of trap operation located in their corresponding stability diagram.

INTRODUCTION

In a previous paper [I], we have shown the possibility of confining ions inside a r.f. quadrupole trap supplied with an impulsional voltage of the form

v, cos (Qt)

1 - k cos (2Qt) (1)

where R/27c is the frequency of the r.f. field, and with 0 d k < 1. The advantage of such a signal, compared with the classic sinusoidal signal

of trap operation, is that it presents periodic zero potential temporal zones during which it is possible to inject ions or electrons inside the trap for collisional studies without changing their initial energy before collisions. In addition, this signal has a narrower frequency spectrum than signals made with suitable square temporal components. So, it can be electronically realised without distortion and remains analytically known during experiment.

We give here the first experimental results of confinement of Xe+ ions in a trap supplied with the voltage defined in eqn. 1. The stability diagram of the trap has been determined and we have compared the efficiency of confinement for equivalent points characterized by the same values of fl,, in the two stability diagrams corresponding to sinusoidal and impulsional excitations of the trap. This comparison was made for fi, = 0.5 (best confinement efficiency)

0168-1176/91/$03.50 0 1991 Elsevier Science Publishers B.V.

Page 2: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

66

and jIZ z 0.2 towards the left boundary of the diagrams (less confinement efficiency) [2].

EXPERIMENTAL ARRANGEMENT AND PROTOCOL

The experimental equipment has been described in a previous paper [3] (Fig. 1). The signal synthesizer generates a periodic voltage of different shapes (Fig. 2) applied to the ring of the trap, while a d.c. voltage - U, is applied to both the end-cap electrodes. In this study, we used a sinusoidal signal, and an impulsional signal given by eqn. 1 with k = 0.8.

The experimental protocol is given in Fig. 3. Xe atoms are ionized inside the trap by electron impact in a vacuum chamber with a total pressure of about 10-s Torr. At the beginning of the Xe+ ion creation, the confinement voltage is applied to the ring electrode. The continement time is predetermined by a computer. When the confinement is turned off, the voltage applied to the ring electrode remains constant with a value which depends on the phase of the voltage at that time. To have a maximum number of detected ions, we must stop the confinement at the maximum voltage amplitude (phase = 0). The two electrodes Dl and D2 maintain a constant electric field between the upper end-cap electrode and the electron multiplier. The profile of the Xe+ ion time-of-flight between the trap and the electron multiplier is given

AMP.

p$za-,

ELECTRON MULTIPLIER -

r-l --mm CONTROL GRIDS J2- - - -

SIGNAL 4lliESlZER

ELECTRON GUN

-&ig&J-l

Fig. 1. Experimental arrangement.

Page 3: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

67

Y$zlf selector

I 1

Comparator

t

I\

contains one Memory - periodofthe

digitalized signal

+lZOV Confinement

Fig. 2. Layout of the signal synthesizer.

by a transient recorder used with an observation window of 1024 temporal channels covering 512 ps. To avoid saturation of the ion detector and space charge perturbation [4], we work with a low density of Xe atoms (~3.5 x 108atomscm-3* m , N lo-’ Torr). In these conditions, we need 2000-3000 cycles of creation-confinement-counting for good visibility and statistics.

Figure 4 shows an example of a temporal signal at the output of the transient recorder under typical experimental conditions. The curve is in two parts: the first (a) corresponds to switching interference; the second (b) gives the time-of-flight profile of the confined Xe+ ions due to the distribution of their velocities and spatial positions when the confinement is turned off. Switching interference does not depend on the confinement time nor on the ion signal amplitude and therefore it can be easily extracted.

Page 4: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

68

One cycle

ELECTRON /L /L GUN ion$reation

/’ // ‘ 1 I

I lTLlON i SIGNAL

I I I ,

TRANSIENT i RECORDER i

I I I

L I I

//

no confinement i I t I I I I I I . . III I I I I

\ I I I /L I 8 /’ 1 4 I I

T.O.F. ion de&ion i

r-i DATA I I /L /L I I PROCESSING /’ //

memorisation accumulation

Fig. 3. Timing protocol.

EXPERIMENTAL AND THEORETICAL STABILITY DIAGRAMS

The motion of an ion in the trap is described by a system of independent Hill equations in each of the perpendicular spatial directions [l]. From the stability diagram in the plane (LL,~) of these equations, we deduce the dia- gram in the plane (U, , V,) by

m z2Q2 u, = - ; y a,(k)

Vm(k) = +E e 2(fy2k) I,(k)

(2)

(3)

Figure 5 shows the diagram A which represents the theoretical stability diagram of the trap for confinement of a Xe+ ion at a frequency s2/27c = 90.9 kHz and for k = 0.8 and z0 = 1 cm; it is compared with the corresponding experimental outlines B-F (k = 0.8) which are obtained for a given lowest amplitude of the signal (defined threshold) which is the same for all the outlines, which only differ by the time of confinement. The smallest value of these times corresponds to the outline B, which defines the largest experimental

Page 5: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

69

NUMBER OF IONS (au.)

+ +*

+ ++ +

+ + +

+ +

+ +

T.O.F. L 0 0 512

Fig, 4. Typical ion signal with accumulation of the results for 2 x 16 cycles: (a), switching interference; (b), ‘time-of-flight profile of ejected Xe+ ions.

area in the plane (U,, VT) and is x 80 ps, i.e. seven periods of the confinement voltage. By contrast, outline F defines values of U, and V,(k) for times of confinement up to several tens of milliseconds. The decreasing efficiency of the trap with time is due to a corresponding increasing of ion losses.

The slight difference between outlines A and B, delimiting respectively the theoretical and experimental stability diagrams, is attributed to the imperfec- tions of the trap and to the ion/ion or ion/atom interactions not taken into account in the theory.

CHARACTERISTICS OF THE CONFINEMENT

We have compared the confinement efficiency of the trap for various points inside the stability diagram. The curves in Fig. 6 show the exponential

Page 6: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

70

_ lJ0 (volts)

0

-2

-4

-6

kEO.8

B to F: experimental

(volts] t ,

30 60

Fig. 5. Stability diagram of the trap in the (U,,V,) plane for Xe+ ions (k = 0.8): theoretical stability diagram (A); experimental stability areas for different confinement times T,: 80 pts (B); 0.8ms (C); 1 ms (D); 5ms (E); and 1Oms (F).

100

50 (a) ,

(b) .

w .

OO 1 I I

3 6 9 12 15

NUMBER OF Xe+ IONS (a.~.)

TC

‘ms)

Fig. 6. Number of confined Xe + ions plotted against the confinement time T, for three operating points with jQO.8) = 27V and U, = OV (a), 0.65V (b) and 1.2V (c).

Page 7: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

71

- ut) (volts)

2

0

-2

-4

-6

theoret ical

Fig. 7. Theoretical stability diagrams of the trap in the (U, , V,) plane for Xe’ ions with the sinusoidal (k = 0) and impulsional (k = 0.8) modes of operation.

decrease of the number of ions versus the confinement time (denoted as T,). The efficiency of confinement decreases when the d.c. voltage U, increases, that is when the confinement point moves away from the centre of the stability diagram.

In the (U,, V,) plane the theoretical stability diagram D(k) of the impulsional mode is wider than the theoretical stability diagram D(0) of the sinusoidal mode and extends to higher values of V, as shown in Fig. 7. Consequently, for two equivalent points, the maximum amplitude V,(k) of the impulsional voltage is always higher than V,(O) of the sinusoidal voltage, which is given from eqn. 3 by

v,(O) = v,(WW)(l - k)

where R(k) = 1,(0)/1,(k).

(4)

Table 1 gives the computed principal operating values relating to equivalent points (k = 0 and k = 0.8) for 8, = 0.479 and /I, = 0.230. The experimental curves given in Fig. 8 for 6, = 0.479 show that the effkiency of confinement is nearly equivalent for sinusoidal and impulsional operation. Figure 9 shows that for the lower value, j?= = 0.230, the efficiency of confinement is better for the impulsional mode of operation. This is a departure from the theoretical predictions because as the efficiency of confinement is an increasing function of the pseudopotential well depth, the results must be identical because

Page 8: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

72

TABLE 1

Operating values of parameters related to two equivalent points located in the optimum confinement zone of the diagrams (/I= = 0.479) and to two equivalent points located near the left boundary of the diagrams (/I, = 0.230), with a, = 0 in each case

B: k I,(k)

0.479 0 0.618 0.479 0.8 0.242

0.230 0 0.330 0.230 0.8 0.129

v,(k)

14.0 27.0

22.4 44

R/2n (kHz)

90.9 90.9

158.7 158.7

equivalent points correspond to the same /I,, that is to the same secular frequency and thus to the same well depth. Investigations are continuing to clarify this point.

CONCLUSION

We have confirmed experimentally the possibility of confining ions in a r.f. quadrupole trap supplied with an impulsional voltage. We have obtained the stability diagrams of the trap for Xe+ ions corresponding to different times of trapping and we have defined a zone of best confinement. Also we have compared the efticiency of the confinement for equivalent points for the sinuoidal and impulsional modes of operation; we notice that for /I, = 0.479 the efficiency is practically the same whereas for /I, = 0.230, that is towards the boundary of the diagrams, the impulsional mode is more efficient. In

NUMBER OF Xe+ IONS (a.u.)

100

50

0

. k = 0.8 . .

k=O ,

I 1 25 50

TC

(ms) Fig. 8. Number of confined Xe+ ions plotted against the confinement time T, for two equival- ent points connected with /IZ = 0.479.

Page 9: Experimental study of a r.f. quadrupole ion trap supplied with a periodic impulsional potential

73

NUMBER OF Xe+ IONS (a.u.)

loo

50

0 I I

0 20 40

TC

P-1

Fig. 9. Number of confined Xe+ ions plotted against the confinement time T, for two equival- ent points connected with & = 0.230.

conclusion, impulsional operation of the trap is at least as efficient as that for the classical sinusoidal mode. Thus it is a suitable mode of trapping when one needs injection of ions or electrons in the trap for collisional studies without a modification of their energy when entering into the trap during the zero potential temporal zones of the impulsional potential.

REFERENCES

1 S.M. Sadat Kiai, J. Andre, Y. Zerega, G. Brincourt and R. Catella, Int. J. Mass Spectrom. Ion Processes, 107 (1991) 191.

2 R. Ifflander and G. Werth, Metrologia, 13 (1977) 167. 3 E.R. Mosburg, Jr., M. Vedel, Y. Zerega, F. Vedel and J. Andre, Int. J. Mass Spectrom. Ion

Processes, 77 (1987) 1. 4 J.E. Fulford, D.N. Hoa, R.J. Hughes, R.E. March, R.F. Bonner and G.J. Wang, J. Vat.

Sci. Technol., 17 (1980) 829.