Ion-Molecule Reactions in the Gas Phase

mation in the Langevin calculation which considers only the ion-induced ...... (2) Coolidge, A. S., James, Η. M., Present, R. D., J. Chem. Phys. 4, 1...
1 downloads 0 Views 2MB Size
7

Energy Transfer in Ion-Molecule Reactions LEWIS FRIEDMAN

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

Brookhaven National Laboratory, Upton, Ν . Y .

11973

Experimental energy transfer processes are studied either by investigating the velocity dependence of certain ion-molecule reactions, the distribution of energy in reaction products, or intramolecular isotope effects. Kinetic energy transfer is observed in competitive dissociation reaction channels in HD-rare gas reactions and the similar dissociation process in the methane system yielding CH3+. In some cases the conjectured mechanism, which requires unit reaction efficiency at every ion­ -molecule collision fails because of the separation of reactant and product potential energy surfaces near possible collision impact parameters. The He -O system demonstrates the importance of considering the nature of the interaction potential. Isotopic studies with He and He show that complex formation in He O reactions provides a mechanism for transferring kinetic energy to the neutral He product. +

3

2

4

+

2

T

his discussion of ion-molecule reactions is limited to processes involv­ ing a chemical change which can be detected by mass analysis of reac­ tion products. Resonant charge transfer between ions and their parent neutral molecules or energy transfer via inelastic collisions will not be included. Emphasis is placed on experimental work done in the C h e m ­ istry Department of Brookhaven National Laboratory which has been directed at testing a relatively simple ion-molecule reaction mechanism. F o r ion-molecule energy transfer studies it is necessary to separate the velocity dependence of the ion-molecule collision cross-section from the velocity or kinetic energy dependence of the ion-molecule reaction crosssection. T h e mechanism proposed by Gioumousis and Stevenson (GS) (8) is particularly attractive because the collision cross-section is calculated directly from the Langevin (25) classical microscopic orbiting cross-section. Gioumousis and Stevenson defined the experimentally observed phenomenological cross-section Q as ^

I nl v

where I is the secondary or product ion current, and I s

v

is the primary or

87

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

88

ION-MOLECULE REACTIONS IN T H EGAS PHASE

reactant ion current observed at the mass spectrometer detector, η is the concentration of neutral molecules, and I is the reactant ion path length in the ion source. (This definition applies only to low pressure reactions where the ratio of I /I is less than 0.05—i.e., where there is a trivial depletion of I in the reaction.) Using a kinetic analysis and as­ suming that reaction takes place at every collision, Gioumousis and Stevenson showed that for ions with a large kinetic energy with respect to the energy of the neutral reactant molecule, Q is given by s

p

p

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

(2) where e is the charge on the ion of mass m μ the reduced mass of the reacting system, a the polarizability of the molecule, and Ε is the electric field in the mass spectrometer ion source. Stevenson and Schissler (23), in a companion paper to Gioumousis and Stevenson's theoretical study, demonstrated that Q's obtained experimentally and from Equation 2 were i n excellent agreement for the reaction of low energy D with D and from the standpoint of data usually obtained i n kinetic studies i n good agreement for a number of ion molecule reactions i n H -rare gas and H -diatomic molecule systems. T h e Gioumouis and Stevenson model is somewhat inadequate for higher velocity ions because of the approxi­ mation i n the Langevin calculation which considers only the ion-induced dipole interaction in the ion molecule potential energy function. H a m i l l and co-workers (1, 14) attempted to account for deviations from the G - S model in reactions of ions having kinetic energy in excess of a few e.v. by including a term in the cross-section expression for hard sphere ion-neutral impacts. T h i s approach, while stimulating, was accepted with reservations because alternative reaction channels, which were not measured in the early experiments, could account for the observed devia­ tions from the theoretical model. h

2

+

2

2

2

Interest at Brookhaven was stirred by the contrast between the excellent agreement between theory and experiment for the D + D reaction and the rather poor description provided for the H - H e and H - N e systems. T h e H - H e system is particularly interesting because of the relatively few particles involved in the reaction and its potential for ac­ curate theoretical treatment. T h e reactions of H or H D + with H e will be among the first to be treated in terms of a theoretically computed potential energy surface; comparison of experiment and theory i n this system is therefore of prime importance. 2

2

+

2

2

2

2

Elemental

+

Systems

Some of the problems encountered in the mass spectrometric study of ion-molecule reactions are illustrated in a review of the H - H e system (25). If the spectrometer ion source is used as a reaction chamber, a mixture of H and H e are subjected to electron impact ionization, and both H and H e are potential reactant ions. T h e initial problem is iden2

2

2

+

+

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

7.

FRIEDMAN

89

Energy Transfer

tifying the reactant ions. Reaction 3 :

Thermochemical considerations suggest that HeH+ +

H

(3)

+ He — HeH+ +

H

(4)

He+ + H

2

is more probable than Reaction 4 H + 2

because the latter is endothermic with ground state H by approximately 1 e.v. while Reaction 3 is strongly exothermic (8.3 e.v.). Identifying the reactant ion in this system is relatively straightforward because of the marked difference in H + and H e ionization potentials and ionization efficiency curves (Figure 1). T h e ionization efficiency curve of H e H 2

+

2

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

+

+

ω ζ 3 >-


°

* 0

I 1 1 1 1.25 2.50 3.75 5.00 AVERAGE TRANSLATIONAL ENERGY-e.v

ι6.25

Figure 7. COH /COD ratio as a function of reactant ion average translational energy in a CO-HD mixture +

+

Accelerating voltage, 2500 volts and ionizing electron current, 10.5 μαπιρ. Closed circles are data taken at 95~.e.v. electron energy with data at 200 e.v. given by the open circles. The smooth curve is the calculated COH /COD ratio. +

+

Intramolecular isotope effects were studied in the systems N -HD, CO-HD, 0 -HD and C0 -HD (20). Product decomposition directly associated with rupture of OH or OD bonds was not observed in these reactions. Isotope effects in decomposition processes which gave OH or OD+ from reactions of 0 + with HD and COH+ or COD+ from 2

2

2

+

2

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

7.

FRIEDMAN

Energy Transfer

99

C 0 - H D reactions were not large enough to shed light on ion-molecule reaction mechanisms. However, the correlation cited above related to energy transfer and product decomposition did hold in these systems. Only minor deviations were observed from the theoretical energy de­ pendence in these ion-molecule reaction cross-sections. Intramolecular isotope effects for the more exothermic processes were almost completely accounted for by the displacement isotope effect. Comparison of calcu­ lated and experimental ratios of A B H + / A B D for C O - H D reactions are given in Figure 7. T h e N H / N D ratios produced in the least exo­ thermic reaction of the set studied were fitted, assuming that 5% of the kinetic energy of N was converted to internal energy and that competi­ tive unimolecular decomposition and displacement isotope effect oc­ curred. T h e A B - H D reactions demonstrate the complexity in identi­ fying the energy distributions on reactant ions since both electronic excited states and vibrational distributions must be considered in identi­ fying potential channels of reaction. T h e available photoionization data and vibronic distributions computed from squares of overlap inte­ grals provide sufficient information to account for the observed reaction cross-sections. 2

+

+

2

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

2

+

+

2

+

Energy Dependence

of

Cross-Sections

R e a c t i o n s o f C o m p l e x I o n s . F o r reactions of systems containing H or H D the failure to observe an E~ dependence of reaction crosssection was probably the result of the failure to include all products of ion-molecule reaction in the calculation of the experimental cross-sections. F o r reactions of complex molecule ions where electron impact ionization probably produces a distribution of vibrationally excited states, kinetic energy transfer can readily open channels which yield products obscured by primary ionization processes. In such cases an E~ dependence of cross-section may be determined; frequently η = 1 has been found. T h e methane system is an interesting example of this problem and is probably typical of many hydrocarbon ion-molecule reactions. Figure 8 shows results obtained in several early investigations (4, 14, 24) of R e ­ action 13. 1/2

2

n

C H + + C H -> C H + + C H 4

4

5

(13)

3

T h i s reaction was considered the only reaction channel because it is the only known channel which is exothermic with ground state C H ions. Reactions yielding C H and C H have been observed and are the least endothermic of the possible reactions of C H with C H . How­ ever, ionization efficiency curves establish C H rather than C H as the reactant ion. Reaction 14: 4

2

5

+

2

4

4

3

CH + + C H 4

4

+

+

CH + + CH 3

+

4

+

+ 4

3

+ H

(14)

2

requires approximately 1.4 e.v. and is difficult to detect because of the rather large yields of C H produced directly from C H in the electron impact ionization. 3

+

4

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

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

100

ION-MOLECULE REACTIONS I N T H E GAS PHASE

Figure 8. Review of data in the literature on the velocity dependence of the reaction C H + Cf/ CH + CH 4

+

+

5

4

S

Solid line gives values calculated for Langevin orbiting cross-sections.

0.5

η

1.0

CONCENTRATION OF MOLECULES

1.5 Χ I0"

13

Figure 9. Dependence of the ratio of C H A ions to to­ tal ions in the methane mass spectrum on the pressure of methane in the mass spectrometer ion source for different values of source repeller voltage +

Recently, the C H - C H reaction has been investigated (9) b y measuring the C H disappearance cross-section rather than C H formation cross-sections. Results of this work are shown in Figure 9. T w o mechanisms cause a loss of C H ions from the total ion yield in the methane mass spectrum. There are loss processes i n the ion source which generate new ions, C H , and possibly other products. Other loss 4

4

+

4

+

5

4

5

+

+

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

+

7.

101

Energy Transfer

FRIEDMAN

mechanisms are those which destroy C H in the mass spectrometer analyzer tube, collision-induced dissociation processes (18), or resonant charge transfer with thermal C H molecules, etc. T u b e and source processes can be separated b y studying the ion repeller or reactant ion energy dependence of total loss processes while holding all other variables constant. D a t a taken in this type of study are presented in Figure 10 along with a set of independently measured C H formation cross-sec­ tions. Extrapolating the plot of C H loss cross-section vs. (eEl)~ gives an intercept which measures the contribution of tube losses. Sub­ tracting this component yields loss cross-sections which are in excellent agreement with the solid line in Figure 10, calculated from the Langevin cross-section for the system C H - C H . The question of which channels account for the difference between the observed C H cross-section and the C H loss is illuminated by studying the isotopic system C H - C D . When mixtures of C H and C D were subjected to electron impact, a pressure dependent yield of C H D was observed which established the reaction mechanism: 4

+

4

5

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

4

+

4

5

+

1/2

+

4

+

+ 4

4

4

4

4

2

CH + + C H 4

4

CH

5

+ CH

+

CH

3

(15)

3

+ H

+

+

2

Here again there is evidence that vibrationally excited C H can react with some kinetic to internal energy transfer and produce ions which + 4

1 i CH; +CH SYSTEM EXPERIMENTAL CH + FORMATION CR0SSSECTION THEORETICAL REACTION CROSS-SECTION 4

5

— 100

EXPERIMENTAL TOTAL ι Δ DISAPPEARANCE CROSS-SECTION > OF CHj X

"

S 80 σ

/

Y

60 If 40 -

/

Φ

/

/

20

/

/

/

/

_

φ φ

-

φ

ι 0.5

1.5 e.v.-

1.0 (ΘΕΙΓ

,/2

, / 2

Figure 10. Comparison of the velocity dependence of the disappear­ ance cross-section of C i f , formation cross-section of CH , and Langevin orbiting collision cross-section, all as a function of recipro­ cal average kinetic energy of ions in the mass spectrometer source 4

+

+

5

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

102

ION-MOLECULE

REACTIONS

I N T H EG A S P H A S E

cannot be formed in isolated collisions of ground state molecule ions and neutral molecules. If the ratio of the observed C H cross-section to the theoretical value for reaction 15 is plotted as a function of energy (Figure 11), this ratio extrapolates back to a value close to unity for reaction of thermal ions. T h e role of internal excitation i n C H is demonstrated in a similar plot for reactions produced from C H , ionized b y impact of 5

+

4

4

1

ι CH++ CH

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007


H e O + + Ο

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

(17)

7.

105

Energy Transfer

FRIEDMAN

HeO+ + e — He + Ο

(18)

could occur, and the dissociative recombination process could give prod­ ucts with enough kinetic energy to permit H e escape. Fite and co­ workers (5) searched for H e O experimentally and did not find it i n mass spectrometer studies of afterglow. T h e possibility of H e O as a tran­ sition species in these experiments was not ruled out. +

+

T h e technique of measuring the 0 produced by reaction of H e existence of H e O .

and 0

+

kinetic energy distribution

showed promise for establishing the

2

Experiments with H e

+

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

+

3

and H e

4

isotopes and 0

were carried out in the ion source of a mass spectrometer.

Retarding

potential curves for Ο in the two systems were determined, and the corn­ +

's

1

1.0

2.0

3.0

RETARDING VOLTAGE, VOLTS

Figure

13.

Retarded ion curves for 0 resulting He - 0 and He - 0 interactions +

z

+

4

2

+

2

from

2

Normalized ion intensities are plotted as a function of retarding volt­ age. The unlabeled curve gives the observed kinetic energy distribu­ tion for reactant *He and He ions (shaded and open squares). A

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

106

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

ponent of Ο , produced by electron impact on 0 , was subtracted from both studies. T h e data obtained from these studies are shown i n Figure 13, which shows a significantly smaller average value of kinetic energy deposited in Ο in reactions of H e with 0 . If the reaction mechanism were one of resonant charge transfer followed by dissociation, the H e and H e isotopes would deposit almost identical amounts of energy in 0 , which could dissociate into Ο and Ο in their respective ground states with 2.93 e.v. kinetic energy in both Ο and O. T h e difference between the 0 kinetic energy distributions obtained with H e and H e pro­ vides strong evidence for the mechanism which proceeds via H e O with H e 0 decomposing and leaving behind a lower velocity 0 . T h e kinetic energy shift observed in the isotopic reactions and magnitudes of the observed energy distributions which correspond to about 1 e.v. mean kinetic energy also support this conclusion. T h i s mechanism produces H e atoms with sufficient kinetic energy to escape the earth's gravitational field. +

2

+

3

2

+ 3

+ 4

+

2

+

+

+ 3

+ 4

+

Downloaded by UNIV OF PITTSBURGH on February 19, 2016 | http://pubs.acs.org Publication Date: January 1, 1967 | doi: 10.1021/ba-1966-0058.ch007

3

Literature

+

+

Cited

(1) Boelrijk, N., H a m i l l , W. H., J. Am. Chem. Soc. 84, 730 (1962). (2) Coolidge, A. S., James, Η . M., Present, R . D., J. Chem. Phys. 4 , 193 (1936). (3) E v e t t , Α . Α . , J. Chem. Phys. 2 4 , 150 (1956). (4) F i e l d , F. H., F r a n k l i n , J. L., L a m p e , F. W . , J. Am. Chem. Soc. 79, 2419 (1957). (5) F i t e , W . L., S m i t h , A. C . M., Stebbings, R . F., J. Geophys. Res. 68, 3225 (1963). (6) F r i e d m a n , L., M o r a n , T. F., J. Chem. Phys. 4 2 , 2624 (1965). (7) Geise, C . F., M a i e r , W . B., J. Chem. Phys. 3 5 , 1913 (1961). (8) Gioumousis, G., Stevenson, D. P., J. Chem. Phys. 29, 294 (1958). (9) G u i d o n i , Α . , F r i e d m a n , L., J. Chem. Phys. i n press. (10) Gustafson, E., L i n d h o l m , E., Arkiv Fysik 1 8 , 219 (1960). (11) H a n s o n , W . B., J. Geophys. Res. 67, 183 (1962). (12) K l e i n , F. S., F r i e d m a n , L., J. Chem. Phys. 4 1 , 1789 (1964). (13) Krauss, M., M i e s , F., private communication. (14) Kubose, D. Α . , H a m i l l , W. H., J. Am. Chem. Soc. 85, 125 (1963). (15) Langevin, P . , Ann. Chim. Phys. 5 , 245 (1905). (16) L i g h t , J. C . , J. Chem. Phys. 4 0 , 3221 (1964). (17) M a c D o n a l d , G . J. F., Rev. Geophysics 1 , 305 (1963). (18) M e l t o n , C . , Rosenstock, H., J. Chem. Phys. 2 6 , 568 (1957). (19) M o r a n , T. F., F r i e d m a n , L. J. Chem. Phys. 39, 2491 (1963). (20) M o r a n , T. F., F r i e d m a n , L., J. Chem. Phys. 4 2 , 2391 (1965). (21) M o r a n , T. F., F r i e d m a n , L., J. Geophys. Res. 70, 4992 (1965). (22) Nicolet, M., J. Geophys. Res. 66, 2263 (1961). (23) Stevenson, D . P . , Schissler, D. O . , J. Chem. Phys. 2 9 , 282 (1958). (24) Stevenson, D . P . , Schissler, D. O ., "The Chemical and Biological Effects of R a d i a t i o n , " M. Haissinsky, ed., p. 249, Academic Press, L o n d o n , 1961. (25) V o n K o c h , H., F r i e d m a n , L., J. Chem. Phys. 3 8 , 1115 (1963). R E C E I V E D A p r i l 29, 1966. Research performed under the auspices of the U . S . A t o m i c Energy Commission.

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