Some Negative Ion Reactions in Simple Gases - ACS Publications

3

Some Negative Ion Reactions in Simple Gases

J. F . P A U L S O N

Downloaded by UNIV QUEENSLAND on April 24, 2013 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch003

Air Force Cambridge Research Laboratories, Office of Aerospace Research, Bedford, Mass.

-

-

Reactions of D withD Oand of O withO ,N O, and NO have been studied with a magnetic sector mass spectrometer. Competition between electron transfer and ion-atom inter change has been observed in the production of O2- by reaction of O withO ,an endothermic reaction. The negative ion of the reacting neutral molecule is formed in O , N O, and NO but not in D O. Rate constants have been estimated as a func tion of repeller potential. 2

2

2

2

-

2

2

2

2

2

A lthough low energy, positive ion-neutral collision reactions have been studied extensively during recent years, little is known of the reactions of even the simplest negative ions. This gap in our knowledge is caused largely by the relatively low negative ion currents available from conventional electron bombardment ion sources, reflecting the low twobody electron attachment cross-sections of most gases. Nevertheless, in weakly ionized gases with predominantly thermal electrons, the loss of electrons by three-body attachment becomes an important process relative to molecular ion-electron recombination at pressures even below 1 torr. Charge transfer and ion-atom interchange reactions of these primary negative ions lead to the creation of different negative (and neutral) species, which, together with the primary negative ions, are ultimately destroyed by such processes as ion-ion recombination and associative detachment. The importance of negative ions in sys tems of chemical interest is most easily assessed if prior knowledge of the rates of these several reactions is available. This paper describes re cent studies of the charge transfer and ion-atom interchange reactions of D ~ and Ο ~ ions with some simple molecules of interest, particularly in the chemistry of the lower ionosphere. Experimental eter.

The reactions were studied with a 6-inch 60°-sector mass spectrom The ion source is of the Nier type (11), machined from a block of 28

In Ion-Molecule Reactions in the Gas Phase; Ausloos, P.; Advances in Chemistry; American Chemical Society: Washington, DC, 1967.

3.

PAULSON

29

Negative Ion Reactions

stainless steel and equipped with narrow electron entrance and ion exit slits so that relatively high source pressures can be maintained. The electron beam, produced by an emission regulated rhenium filament, enters the source through a 1.17 X 0 . 1 5 m m . slit, and except for some of the experiments involving D ~ ions, is collimated by a 200-gauss mag netic field. T h e beam passes midway between the repeller plate and the ion exit slit and impinges on a trap maintained at the source potential. T h e electron current measured with a positive bias on the trap is 1 X 10 ~ A . T h e repeller forms one side of the source chamber and is 4.0 m m from the exit slit. Gas enters the source behind the repeller plate and must collide many times with the walls before reaching the region of the electron beam. T h e source block is maintained at 200° ± 1 0 ° C . Ions exit from the source through a 7.15 X 0.114-mm. slit and are accelerated by a potential of 3500 volts. T h e ion-accelerating region is evacuated with a 2-inch mercury diffusion pumping system which maintains this region at a pressure two orders of magnitude below that i n the source chamber itself. T h e analyzer is evacuated with a 6-inch mercury diffu sion pumping system, maintaining a pressure below 1 X 10 ~ torr when the pressure in the source is as high as 0.2 torr. Ion currents are meas ured with an electron multiplier whose gain has been calibrated with the aid of a Faraday cup for all species reported here except N 0 ~ , for which the gain is assumed to be the same as for N 0 ~.


7

6

2

2

Pressure in the source is measured with a micromanometer con nected to the gas inlet line following a variable leak. T h e microma nometer has been calibrated against molecular number densities in the source determined by collecting on the repeller all positive ions produced in each of several gases of known ionization cross-section (14) with an electron beam of known energy and intensity. These calibrations agree to within 12% with pressure measurements made by connecting the micromanometer directly into the source chamber itself. T h e mass spectrometer is equipped with an X Y recorder and a motordriven potentiometer on the ionizing voltage control, giving continuous traces of ion current as a function of electron energy. Continuous traces are also obtained for ion current as a function of pressure. F o r the stud ies described here, in which low electron energies were used, we had to adjust the nominal ionizing voltage as the repeller potential was varied so that the desired electron energy could be maintained. Calibration curves obtained by observing the shift in the optimum nominal electron energy for production of Ο ~ from C 0 and in the half-width (full width at half maximum) of that peak are shown in Figures 1 and 2. T h e electron energy scale has been normalized at a repeller potential of 1 volt to the electron energy, (8.2 e.v.), which Schulz (18) found to be optimum for C 0 using a quasi-monochromatic electron beam. T h e normalization involved increasing the nominal electron energy, read on a differential voltmeter, by 0.2 e.v. T h e optimum electron energy for production of the O peak, whose onset was found by Schulz (18) to be 3.85 ± 0.1 e.v., is then 4.3 e.v. compared with Schulz's value of 4.4 e.v. and with 4.3 e.v. obtained by R a p p and Briglia (13). T h e difference in optimum nominal electron energies for the two processes by which Ο ~ is formed in C 0 is independent of repeller potential (Figure 1). T h e half-width increases from 1.6 e.v. at 1 volt to 3.4 e.v. at 16 volts repeller potential. 2

2

-

2



30

ION-MOLECULE REACTIONS IN T H E GAS PHASE

REPELLER POTENTIAL (VOLTS) Figure 1.

Calibration curves for electron energy at different repeller potentials

< x

i.o I ο

1 2

1 « 1 1 1 1 4 6 8 10 12 14 REPELLER POTENTIAL (VOLTS)

ί

16

Figure 2. Half-width of the 0 ~ peak from C 0 maximizing at an electron energy of 8.2 e.v., at different repeller potentials 2

Electron energies reported here are those obtained from the nominal values after correction with calibration curves shown in Figure 1. Results

In this work the reacting negative ions were produced with electrons of low energy leading to resonance capture processes, rather than at higher



3.

31


PAULSON

energies where secondary electron capture and ion pair production occur. T h i s method has the advantage that one ion species can be produced and its reactions studied largely to the exclusion of other processes possibly leading to the same product ion. In order to obtain easily measurable ion currents, however, it was necessary to use electron energies above the capture threshold—a procedure which results in the products of the dissociative resonance capture process having kinetic energies above that of the ambient gas. We assume that there is no spatial anisotropy in the velocity distributions of these products. Initially then, negative ions produced at any point in the electron beam have velocity vectors whose envelope is a sphere. If the repeller potential is sufficiently high, ions initially directed away from the plane of the exit slit are decelerated, their trajectories reversed, and finally accelerated toward the plane of the slit. T h e ion residence time and path length in the source may be estimated if one assumes that the ion exit slit is infinitely thin and that the only ions collected are those whose initial motion is directed per pendicular to or away from the plane of the slit. T h e average residence time, r, for these ions is: d

- {2E e

v

T =

t

r

+

V)

(1)

+

V)

(2)

r

and the average path length, λ is: λ

-

d

(2E

t

r

In Equations 1 and 2, d is the distance from repeller to exit slit, V is the repeller potential, E is the initial ion kinetic energy, and m/e is the massto-charge ratio of the ion. Exact expressions for τ and λ for all the pri mary ions collected cannot be derived without information on the collec tion efficiencies for ions produced in the various regions of the source chamber, information not available for the type of source used here. Although data were taken over a range of repeller potentials, uncertain ties i n average residence times and in mass discrimination effects are more serious at low than at high V , and rate constants are therefore more reliable under the latter condition. A t repeller potentials high compared with the initial kinetic energy of the reacting ions, Equation 1 reduces to the usual expression for residence time of an initially thermal ion in a uniform electric field. T

t

T

Discussion D~~ + D 0 . Figure 3 shows the ion currents of D ~ , 0~, and OD~~ observed in D 0 as a function of electron energy. Neither D 0 ~ nor H 0 ~ was observed. T h e close similarity between the curves shown for D ~ and O D ~ suggests that the exothermic reaction 2

2

2

2

D - + D 0 -> O D - + D 2

occurs.

2

T h i s reaction has been suggested by Muschlitz (10).


(3) T h e re-

32

I O N - M O L E C U L E REACTIONS IN T H E GAS P H A S E

suits of varying the pressure of D 0 are shown in Figure 4, where the currents of D and of O D - vary with the first and second power, respec tively, of pressure. T h e current of Ο ~, not shown, varied linearly with pressure. T h e electron energy used to produce the D ~ ions, whose varia tion with pressure is shown in Figure 4, was 6.4 e.v.—i.e., the optimum energy shown in Figure 3. T h e initial kinetic energy with which these D ~ ions are produced may be estimated from values of the electron affi nity of D ~ (assumed to be the same as that of Η (12) ) and the bond dis sociation energy of D 0 (assumed to be the same as that of H 0 (8)) since, from conservation of energy and momentum: 2

_

_

2

E


t

2

= (1 -

β) [E

e

-

A)]

(D -

(4)

In Equation 4 E is the kinetic energy of the fragment ion; β = m / M where m, is the mass of the fragment ion, and M is the mass of the target molecule; E is the electron energy; D is the bond dissociation energy; and A is the electron affinity. A p p l y i n g Equation 4 to the present case gives a value of E = 1.9 e.v. T h e residence time from Equation 1 is 2.14 X 10 ~ sec. at 10 W . Rate constants for Reaction 3 derived from the relation: t

? :

0

e

t

7

0

5

10

15

ELECTRON ENERGY (e.v,CORRECTED) Figure 3. Ion currents of D~, Ο ~, and OD~ from D 0 as functions of electron energy. The curves are traced from XY recordings and do not show relative in tensities 2


0


3.

PAULSON


33

.6 .8 I 2 4 6 8 10 IO" xn Jem." ) 2 l 4

3

D

Figure 4. Ion currents of D~ and OD ~ from D 0 as functions of D 0 number density, using an electron energy of 6.4 e.v. 2

2

are shown i n Table I. In Equation 5, i and i are the ion currents of reactant and product ions, respectively, and n is the number density of reactant neutral species. Although mass discrimination effects are particularly troublesome for such low mass ions as D ~ and may lead to erroneously large rate constants, it appears that Reaction 3 is very rapid, occurring essentially at every collision. x

2

0

T a b l e I. Repeller Potential V r

4 5 6 7 8 9 10 11 12 13 14 15

Rate Constants

(cc. molecule' Reaction Reaction 6 14

1

Reaction 3 454 466 463 468 457 433 380 375 364

1.7 2.3 3.0 4.3 5.6 6.4 7.6 8.7 9.3 10.8 10.9 12.1

3.4 3.6 3.7 3.5 3.7 3.6 3.6 3.5 3.6

sec.' ) X 10 Reaction Reaction 18 20 1

0.20 0.23 0.24 0.27 0.34 0.44 0.55 0.63 0.74

11

103 119 130 128 130 122 119 131 143


Reaction 22 1.8 2.0 2.0 2.1 2.3 2.4 2.7 2.9 3.3

34


O" + 0 . 2

T h e reaction: Ο-

+ 0

2

-> 0 - + Ο

(6)

2

was suggested by Burch and Geballe (2) to explain their data from a pulsed Townsend discharge i n oxygen. Although Reaction 6 is endothermic by about 1 e.v., the production of 0 ~" was observed in this work even at low repeller potentials probably because of the excess kinetic energy of O ions produced by dissociative electron attachment to 0 at electron energies above threshold. Figure 5 shows the data on Ο ~ and 0 ion currents with varying electron energy. T h e curves maximize at electron energies differing by about 1 e.v., reflecting the fact that R e 2

-


2

2

-

0 5 10 ELECTRON ENERGY (e.V,

15 CORRECTE D)

Figure 5. Ion currents of 0~ and Or from 0 as functions of electron energy. Traced from XY recordings 2

action 6 is endothermic and that Ο ~ ions i n the high energy end of the distribution react much more rapidly than those with lower kinetic energy. T h e possibility that 0 ~ might be formed by the three-body attachment process 2

e + 0

2

+ 0

2

-> 0 - + 2

0

2

as suggested by Muschlitz (10) rather than by Reaction 6, cannot be unequivocally eliminated since the same second power dependence of 0 ~ ion current upon sample pressure (Figure 6) would be observed i n either case. However, it is extremely improbable that the optimum electron 2


3.


PAULSON

35

energy for the three-body attachment process would be as high as shown in Figure 5—i.e., 6.9 e.v. T h e initial kinetic energy of 0 ~ ions produced by dissociative attachment in 0 at an electron energy of 6.9 e.v. may be determined from Equation 4 to be 1.64 e.v. using values of 1.465 e.v. (2) for A(0) and 5.09 e.v. (7) for D(0—O). T h e residence time for 0 ~ ions calculated from Equation 1 is 6.0 X 10 ~ sec. at 10 volts repeller potential. Rate constants for Reaction 6 determined from data at varying V are shown in Table I and are seen to increase sharply with increasing repeller potential, as expected for an endothermic process. 2

7


r

Figure 6. Ion currents of 0~ and 0 ~ from O2 as functions of 0 number density, using an electron energy of 6.9 e.v. 2

2

Reaction 6 might conceivably proceed through either an ion-atom interchange mechanism (Reaction 7) or a charge transfer mechanism (Reaction 8), 16Q-

+

18Q

16Q-

+

iSQ

n

2

_^

160

18Q- +

18Q

ISO"

Using an equimolar sample mixture of O and l f

2

+ 1 8

180

(7)

0

(8)

1 6

0 , together with a small 2


36


amount of 0 0 impurity, the following set of reactions contributing to the ion-atom interchange mechanism must be considered: 1 6

1 8

i6Q -

+

i8Q-

+

ICQ"

1 6

0

+

180

18Q- +

160

16Q- - f 160 16Q- + 18Q- +

1 6

1 6

0

0

0

1 8

0

0~

+

1 8

0

160

1 8

0"

+

1 6

0

0

1 6

1 6

—

1 6

1 8

0~

1 8

0 -

+

0~

1 8

0 -

1 8

+

0 1 8

+

2

0

1 6

+

2

0

0

1 8

1 6

+

2

1 8

0

0

+

2

0 -

160 ->

1 8

1 8

1 6

2

0

0 -

1 8

t

2

1 8

0

18Q- +

1 6

2

i80

1 6

0

0


The following set contribute to the charge transfer mechanism: 16Q- + 16 -

+

180ISO"

+ +

0

16Q -

+

1 6

18Q- +

1 8

0

1 8

2

160

0 18θ

1 6

0

+

2

1 6

+

?

2

160 0 18

0

0 180

1 8

2

2

0

0

1 6

0 - + 180 - +

1 6

2

1 8

0 -

160 -

2

1 6

0

1 8

0~

+

1 6

0

1 6

0

1 8

0~

+

1 8

0

Realizing that the last four reactions of the ion-atom interchange mech anism listed each have only one-half the statistical probability of occurring as do the first four and assuming no isotope effect on the rate constants, we can write the following set of rate equations: —J—

= ^16^36 +

kiii ri32 + S

dt ^7

dt

= kii ri32

di —- = dt

u

~ kiii&tlu

+

^ k}hs1lM +

2

2

+

\ kiUtfiu + 2

^2*16^34 + &2*18^34

kii^n^

k i\%nz2 + 2

(9)

(10)

1 ^ll*18^36 +

- ^1^18^34 + #2*16^36 + &2* 18^36 A

(H)

In Equations 9, 10, and 11, ij is the number density of the ionic species of m/e = y, and n^, n , and n e are the number densities of 0 , 0 0 , and 0 , respectively, in the ion source. T h e rate constants are k and k for the ion-atom interchange and charge transfer mechanisms, respectively. Assuming equal collection efficiencies for all ionic species, we found experimentally that z*i « lis, n » n^, and n 4 = 4.3 X 10~~ n 2 in the sample mixture. Realizing that at t = Ο, ί = = = 0, and that the consumption of reactant ions is negligible, we can write from Equation 9, u

1 8

1 6

3

1 6

2

1 8

x

2

2

6

Z2

2

3

3

Ά2

τ- = 2.043 kxTinr + 0.086 & η τ 2

32

(12)

£l6

and from E q u a t i o n 10 ~

= 1.0215 k nz2T + 2k n, r x

2

2

Ϊ16


(13)

3.

37


PAULSON

For simplicity the residence times, r, of 0 ~ and of 0 ~ are assumed to be equal here. W i t h Equations 12 and 13 it is now possible to deter mine the rate constants k and k at different V . T h e results are shown in Table II and agree fairly well with the rate constants obtained using ordinary oxygen. T h e ratios of the rate constants for the two mechan isms are constant in the range of repeller potentials from 7 to 12 volts. 1 6

x

T a b l e II.

1 8

2

r

R a t e Constants» for 0~

+ 0

2

-> 0 "

+

2

Ο

Repeller Potential, V%


I o n - A t o m Interchange Charge Transfer a

4

5

6

7

8

9

10

11

12

0.7 2.5

0.9 3.1

1.2 3.9

1.5 4.4

1.8 5.1

1.9 5.5

2.1 5.9

2.2 6.4

2.2 6.6

Rate constants given as: k (cc. molecule

-1

sec. ) X -1

10

11

O " + N 0. T h e high pressure negative ion mass spectrum of N 0 contains peaks at m/e 16, 30, 32, 44, and 46 in addition, of course, to the corresponding isotope peaks. T h e currents of m/e 16 and m/e 30 as functions of electron energy are shown in Figure 7. T h e data show only a small shoulder instead of a well defined peak in the current of 0~ at electron energies below 1 e.v., as obtained by Schulz (18) and by Curran and Fox (5) using a quasi-monochromatic electron beam. T h e electron current falls away rapidly in our source at energies below 1 e.v., and we have not attempted to correct the data for this effect. 2

2

0

5

10

15

ELECTRON ENERGY (e.v.CORRECTED) Figure 7. Ion currents of 0~ and NO ~ from N 0 as functions of electron energy. Traced from XY recordings 2


38


T h e dependence of the currents of m/e 16 and m/e 30 upon sample pressure, using an electron energy of 2.3 e.v., is shown i n Figure 8. T h e linear variation of m/e 16 and the quadratic variation of m/e 30 with pressure, together with the results shown i n Figure 7, indicate the oc currence of Reaction 14. O - + N 0 -> N O - + N O

(14)

2

Reaction 14 is 1.0 e.v. exothermic. Using equation 4, with E = 2.3 e.v., JD(N —O) = 1.68 e.v. (8), and A ( O ) = 1.465 e.v. (1), the initial kinetic energy of the Ο ~ ions is calculated as 1.34 e.v. However, Schulz (19) has shown that the kinetic energy of Ο ions from N ? 0 is independent, of electron energy in the range 1.5-3 e.v. and has the value 0.65 e.v. T h i s unusual behavior is ascribed (19) to increasing vibrational excita tion of the N fragment with increasing electron energy. Such behavior is possible i n principle whenever an atomic ion is produced from a t r i atomic or larger parent. Using a value of E = 0.65 e.v. (19), rate constants for Reaction 14 have been calculated and are shown in Table I. A sample of N N O was used to investigate whether Reaction 14 proceeds through a dissociative charge transfer mechanism involving an e

2


-

2

t

1 4

1 5

100 80 60 40 m 30

5 20 xi

σ σ> I 0 £ 8

Ο

2

I .6 .8 i 2 3 4 6 8 10 Ι0" χΠ (cm.- ) Ι 4

3

Ν

Q

Figure 8. Ion currents of Ο" and ΝΟ~~ from Ν 0 as functions of N 0 number den sity, using an electron energy of 2.3 e.v. 2

2


3.

unstable intermediate (Reaction 16). O- +

i 4

N

i 5

39


PAULSON

N

Q

(Reaction 15)

-> ( 0 N N O ) ~ 1 4

1 5

1 4

or by abstraction of Ν

NO~ +

1 5

NO

1 4

N

1 5

NO

1 4

NO~ +

1 5

0~

and 1 5

O- +

by

NO- +

1 4

NO

(15)

NO

(16)

T h e ratio of negative ion currents, z ]A*3o, was found to be 0.80 ± 0.05. Although production of N O ~ might occur partly by abstraction of the N rather than the N as shown in Reaction 16, this mechanism appears less probable than that indicated. We tentatively conclude that Reac tion 15 predominates. Reaction 14 has been proposed by others workers (3, 15) to account for the N O ~ observed in N 0 . It has also been suggested (15) that N O ~ is formed in low abundance by dissociative attachment in N 0 at an electron energy of 25 e.v. Using values of A (NO) = 0.9 e.v. (6) and D(N—NO) = 4.72 e.v. (8) the minimum electron energy for this process is 4.07 e.v. In the experiments described here, a nominal electron energy of 2.9 e.v. was used. Although the energy spread is sufficiently large that some N O ~ might be formed by the dissociative attachment reaction, our data do not indicate a process linear with pressure, producing N O ~ at 2.9 e.v. in the pressure range investigated. A s mentioned above, negative ion currents at m/e 32, 44, and 46 were observed. T h e current at m/e 32 varied with the second power of the sample pressure and showed the same dependence upon electron energy as the m/e 16 current. T h e reaction: 3

1 5


J 5

1 4

2

2

O - + N 0 -> 0 2

+ N

2

(17)

2

which has been suggested by Burtt and Henis (3), is exothermic by 2.4 e.v. T h e ion current ratio i^/iu increased rapidly with increasing re peller potential. T h e rate constants obtained in this work for Reaction 6 permit one to estimate the impurity level of 0 required to produce the observed ion current of 0 ~ . T h e result is that an 0 impurity level of 1% would be sufficient. T h e observed 0 impurity level in the positive ion spectrum is 0.8%. Therefore, the O ~ observed here probably arises from Reaction 6 rather than Reaction 17. T h e ion currents at m/e 44 and 46 were also directly related to the current at m/e 16 in their dependence upon electron energy. In a sample of N N O these peaks shifted to m/e 45 and m/e 47, respectively, and are therefore ascribed to N 0 ~ and N 0 ~ . A n approximately secondpower pressure dependence of the ion current of N 0 ~ was observed both in samples of N 0 alone and in an equimolar mixture of N 0 and 0 , and Reaction 18 is therefore suggested. 2

2

2

2

s

1 4

1 5

?

2

2

2

2

2

Ο- + N 0 2

N 02

+ Ο

(18)

Rate constants for this process are given in Table I. T h e electron affinity of N 0 is unknown. However the results indicate that the reaction is 2


40


endothermic, and therefore the affinity of the state produced is less than that of 0 T h e origin of the N 0 observed is uncertain. Reaction 19, pro posed by B u r t t and Henis (3), is calculated to be exothermic by at least 0.5 e.v., assuming the electron affinity of N 0 is >3.8 e.v. (4). 2

-

2

Ο- + N 0 — N0 2

+ Ν

2

(19)

Our data give a rate constant of 4 X 10 cc. molecule s e c , assuming the production of N 0 solely by this reaction. However, one must consider the probability that N 0 , present as a minor impurity or pro duced by pyrolysis or N 0 of or near the hot filament, would react by Reaction 20. - 1 3

2

- 1

- 1

-

2


2

Ο- + N0

N0 -

2

+ Ο

2

(20)

Using a rate constant for the latter reaction of 1 X 1 0 cc. m o l e c u l e sec. (see below), it is estimated that the N 0 current observed could be explained by an N 0 impurity level of 0.2%. In an equimolar mixture of N N O and 0 run at an electron energy of 5.9 e.v.—i.e., the optimum energy for production of O from 0 , the negative ion current ratio, iis/iu, was found to be 4.06. The predominant negative ion above m/e 44 in this sample was m/e 45, N N 0 , indicating that Reaction 18 proceeds principally through a charge transfer mechanism. Ions observed at higher m/e are of am biguous composition and do not permit elucidation of the origin of the N0 observed. T h e positive ion spectrum showed no isotope exchange occurring in the sample mixture. Ο" + N 0 . Production of N 0 by Reaction 20 - 9

- 1

2

-1

-

2

1 4

1 5

l e

1 8

2

-

2

1 4

1 5

2

l f ,

-

-

2

2

O

-

+ N0

2

-

-> N 0

2

+ Ο

-

(20)

has been observed previously {4, 7), and its cross-section was estimated (7) to be 2.9 times larger than that of the positive ion-molecule reac tion H,0

+

+ H 0 — H 0+ + OH 2

(21)

3

at a repeller field of 4 volts c m . Although the rate constant and energy dependence of Reaction 21 are the subjects of some question (16, 17, 20), the calculated rate constant for Reaction 20 is approximately 1 X 1 0 cc. m o l e c u l e s e c . at low ion energies on the basis of these earlier results. Figure 9 shows the ion currents of O , N O , 0 and N 0 " as functions of electron energy observed in a sample of N 0 at a number density of 3 X 10 c c . . T h e ion currents of both m/e 16 and m/e 30 varied linearly with sample pressure when the respective optimum elec tron energies shown in Figure 9 were used. T h e ion current of 0 , using the lower of the two optimum electron energies shown i n Figure 9 (i.e., 2.2 e.v.) varied as the second power of the sample pressure whereas that at 5.2 e.v. varied linearly with pressure. T h e ion current of N 0 " at an - 1

- 9

-1

- 1

-

-

2

-

2

2

14

- I

-

2

2


3.

PAULSON


41

electron energy of 2.2 e.v. varied as the square of the sample pres sure. These results are consistent with Reactions 20, 22, and 23. O- + N0

— 0 -

2

e + N0

?

2

+ NO

(22)

+ Ν

(23)

0 2

T h e 0 impurity level was well below that required to produce the ob served 0 ~ by Reaction 6. Using Equation 4 and the values D ( N O — O ) = 3.11 e.v. (8) and A(O) = 1.465 e.v. (2), the kinetic energy of O " ions produced from N 0 at an electron energy of 2.2 e.v. is 0.36 e.v. Rate constants for Reactions 20 and 22 calculated from the data at varying repeller potentials are shown in Table I. Attempts to elucidate the mechanisms of these reactions by using a sample of NO? and 0 were unsuccessful because isotope exchange occurred in the sample reser voir. W i t h increasing repeller potential there is increasing electron energy band width and therefore increasing overlap between the 0 ~ 2

2

2


1 8

2

2

0 5 10 15 ELECTRON ENERGY U N C O R R E C T E D ) Figure 9. Ion currents of 0~, NO", 0 ~, and i V 0 ~ from N0 as func tions of electron energy. Traced from XY recordings 2

2

2

peaks produced by Reactions 22 and 23. T h e data for Reaction 22 (Ta ble I) have not been corrected for this effect. Corrections based on the variation of half-widths of peaks with varying repeller potential, shown in Figure 2, indicate that the estimated rate constants for Reaction 22 are independent of repeller potential.


42


T h e production of N 0 ~ b y charge transfer from either S F ~ or S F ~ , previously observed by Curran (4), has been confirmed during this work. T h e attachment cross-section curves for S F ~ and for S F ~ closely overlapped i n the low energy range, and it was impossible to de termine whether S F ~ or S F ~ was the reacting ion. 2

6

5

6

6

5

5


Conclusion T h e work described here is preliminary i n the sense that the rate constants obtained are averages over a wide range of interaction ener gies and may not apply to ions having well defined kinetic energies. T h e assumption of isotropic spatial distributions of the reactant and product ions may lead to seriously erroneous rate constants at low repeller poten tials but becomes unnecessary at high repeller potentials. Similarly, the expression used to estimate ion residence times becomes more accurate at high repeller potentials. Finally, a high field strength ensures that anisotropic reactive scattering effects are minimized. Unfortunately, the width of the reactant ion kinetic energy distribution increases with increasing field strength, and defocusing of the ion beam also occurs, leading to decreased signal levels. However, except for Reaction 6, known to be endothermic, and Reaction 18, the rate constants obtained here show only a weak dependence upon repeller potential, the effect observed most probably being caused largely by a slowly varying ion collection efficiency. Nevertheless, it is clear that definitive tests of the dependence of these rate constants upon kinetic energy can only be car ried out using ion beam techniques in which angular distributions of the products are measured. Acknowledgements T h e assistance of F . Dale and J . Welsh i n performing the work described in this paper is gratefully acknowledged. Literature

Cited

(1) Branscomb, L . M . , Burch, D. S., Smith, S. J., Geltman, S., Phys. Rev. 111, 504 (1958). (2) Burch, D. S., Geballe, R., Phys. Rev. 106, 188 (1957). (3) Burtt, B. P., Henis, J., J. Chem. Phys. 41, 1510 (1964). (4) Curran, R. K., Phys. Rev. 125, 910 (1962). (5) Curran, R. K., Fox, R. E . , J. Chem. Phys. 34, 1590 (1961). (6) Farragher, A. L., Page, F. M . , Wheeler, R. C., Discuss. Faraday Soc. 37, 203 (1964). (7) Henglein, Α., Muccini, G. Α., "Chemical Effects of Nuclear Transforma tions," pp. 89-98, International Atomic Energy Agency, Vienna, 1961. (8) McBride, B. J., Heimel, S., Ehlers, J. G., Gordon, S., "Thermodynamic Prop erties to 6000°K. for 210 Substances Involving the First 18 Elements," NASA Scientific Publication SP-3001, Lewis Research Center, Cleveland, 1963. (9) Moiseiwitsch, B. L., "Advances in Atomic and Molecular Physics," D. R. Bates ed., p. 61, Academic Press, New York, 1965. (10) Muschlitz, Ε. E . , J. Appl. Phys. 28, 1414 (1957). (11) Nier, A. O., Rev. Sci. Instr. 18, 398 (1947). (12) Pekeris, C. L., Phys. Rev. 126, 1470 (1962).


3. (13) (14) (15) (16) (17) (18) (19) (20)

PAULSON


43

Rapp, D., Briglia, D. D., J. Chem. Phys. 43, 1480 (1965). Rapp, D., Englander-Golden, P., J. Chem. Phys. 43, 1464 (1965). Rudolph, P. S., Melton, C. E., Begun, G. M., J. Chem. Phys. 30, 588 (1959). Ryan, K. R., Futrell, J. H., J. Chem. Phys. 42, 824 (1965). Ryan, K. R., Futrell, J. H., J. Chem. Phys. 43, 3009 (1965). Schulz, G. J., Phys. Rev. 128, 178 (1962). Schulz, G.J.,J. Chem. Phys. 34, 1778 (1961). Tal'roze, V. L., Frankevich, E . L., Zh. Fiz. Khim, 34, 2709 (1960). 1966.


RECEIVED May 13,


Some Negative Ion Reactions in Simple Gases - ACS Publications

Recommend Documents