Determination of the equivalence point of the atomic nitrogen+

May 1, 1974 - Determination of the equivalence point of the atomic nitrogen + mononitrogen monoxide titration reaction by electrical conduction...
0 downloads 0 Views 1MB Size
S. E. Schwartz and S. M. Butler

on the (1OiO) face will be delayed and take place rather in one rush at relatively higher temperatures.

Supp1enaentar-y Material Available. Table I will appear following 1 hese pages in the microfilm edition of this volume of the jownal. Photocopies of the supplementary material from this paper only or microfiche (105 x 148 mm, 2 4 reduction, ~ negatives) containing all of the supplementary material for the papers in this issue may be obtained from the Journals Department, American Chemical Society, 1155 16th St., N.W., Washington, D. C. 20036. Remit check or money order for $3.00 for photocopy or $2.00 for inicrofiiche, referring to code number JPC-741116. References 8 urd Neltes T Morimo < , M Nagao and F Tokuda, Buli Chem SOC Jap 41, 1533 (1968) (2) T Morimolo M Nagao and M Hirata, Kolloid-Z 2 Polym 225, 29 (1968) (3) M Nagao dnd T Worimoto, J Phys Chem 73, 3809 (1969) (')

(4) T. Morirnoto and M . Nagao, Bull. Chem. Soc Jap.. 43, 3746 (1970). (5) M. Nagao, J. Phys. Chem.. 75,3622 (1971). (6) B. E. Fisher and W. G. McMilian, J. Amer. Chern SOC., 79, 2969 (1957), (7) S . Rossand H . Clark, J. Amer. Chem. S O C . ,76, 4291 (1954). (8) S. Ross, J. P. Olivier, and J J. Hinchen, Adiian. Chem. Ser., No. 33,317 (1961). (9) J. P. Oiivier and S. Ross, Proc. Roy. Sac.. Ser. A . 265, 447 (1962). (10) W. D. Machin and S. Ross, Proc. Roy. Soc.. Ser. A . 265, 455 (1962). (11) T, Morimoto, K. Shiomi, and H,Tanaka, Bull. Chem. Soc Jap.. 37, 392 (1964), (12) T. Morimoto, M. Nagao, and F Tokuda. ,I. Phys. Chem., 73. 243 (1969), (13) T. Morimoto, M. Nagao, and J. limai, 8uY. Chem. Soc. Jap.. 44, 1282 (1971). (14) See paragraph at end of text regarding supplementary material. (15) S. Dana and W. E. Ford, "A Textbook of Mineralogy," Wiley, New York, N, Y., 1960. p 480. (16) A . L. Dent and R. J. Kokes, J. Phys. Chem . '13, 3781 ('1969) (17) A . Atherton, G. Newboid, and J. A. Hockey Discuss. Faraday SOC.. 52, 33 (1971). (18) W. R. Smith and 0. G. Ford, J. Phys. Chem.. 69, 3587 (1965). (19) €3. W. Davisand C. Pierce, J. Phys, Chem.. 10. 7051 (1966). (20) P. T. Dawson, J. Phys. Chem.. 71, 836 (1967). (21) H, F. Holmes, E. L. Fuller, Jr,, and 6 . H. Secoy. J. Phys. Chem.. 72, 2293 (1968).

ion of the Equivalence Point of the N 4- NO Titration Reaction by

onductio I

Schwartz" and S. M. Butler

Department of Chemistry, State University of New York. Stony Brook. New York 1 7 790 (Received January 24. 7974) PiJblicEIfion costs assisted by the Petroleum Research Fund and the State University of New York

Electrical conduction has been measured in mixtures of atomic nitrogen and oxygen in formed by allowing NO to react with N in a flowing afterglow. Ionic species may be extracted by electric fields of - 5 Vjcm, established between parallel stainless steel electrodes, a t pressures of 60 to 1300 Pa. Any remaining conduction, measured subsequent to extraction, vanishes sharply at the equivalence point of the tiNO N2 0 (1) and is thus attributed to ionic species formed by reaction of tration reaction N atomic nitrogen. The rate of ionization exhibits a first-order dependence on [N] for [O]/[l%'] > 100, and this first.order contribution to the ionization rate exceeds higher order contributions for [OJ/[N] > 10. Sevcxal mechanisms for chemiionization are considered; such ionization may take place by the reaction O ( % ) N(4S)--* NO+ + e-. Measurement of saturation current is found to serve as a sensitive indicator of the equivalence point of (1).

+

-

+

+

Introduction The rapid reaction

rcrc4ss)-I- NO

-

N~

+ o(3p)

(1)

is commonly used i n flowing afterglow systems as a titration reaction for atomic nitrogen,l-ll and as a source of atomic oxyge!n.12,13Typically molecular nitrogen is partially dissociated in an electrical (often microwave) discharge, and KO i.s added to the flowing N2-N mixture downstream from the discharge. Successful application of the technique requires precise determination of the NO flow equal to the N atom flow. This "equivalence point" is most frequently determined by monitoring molecular electronic chemiluminescence present under conditions of ~ means of deterinsufficient or excess NO f l o ~ . I -Other mining this equivali:nce point have included e ~ r , vacu~ - ~ The Journal of Physical Chemistry. Vol. 78. No. 1 7 . 1974

um uv absorption,? and mass-spectrometric measurements of neutrals and ionics concentrations. The validity of the equivalence point determined hy electronic luminescence has been confirmed by these several techniques, from the total pressure decrease accompanying reaction 1,loand from mass-spectrometric product analysrs.ll This work has been recently reviewed.lh2l5 Because of its sensitivity and convenience, measurement of electronic luminescence remains favored as an equivalence point indicator for reaction I. The details of the several mechanisms responsible for these lummescence systems are complex, incorporaling highly specific populating and depopulating processes. Nevertheless, it is possible to describe the intensities of these emission systems, at constant pressures in the range of 100 to 1000 Pa and at low atom concentrations, as proportional to the concentrations of the reactant species3~1*.1G23

Equivatence Point ~ e t e r i ~ i n a tby i ~ Electrical n Conduction P;(~s) -I-N(%) --*

-p O(3P)

~ ~ ( 4 s )

--+

M,*

-

NO*

--,

hr(yeiicw) h~(uv,blue)

Iy a [N]' I,,

3:

1121

(2)

[N][O] (3)

q 3 P ) .+ iZi0 -*- K'C)," hutgreen) I, [01[NOl (4) eactions 3 and 4 are utilized as indicators for the equivance point of (I), since the intensities I,, and Ig exhibit a first-order dependence on the degree of under- or overtively, the concentration of atomic oxygen. ively constant near the equivalence point. which decreases quadratically as the equivalence point is a~proached,is not a sensitive indicator, since the ratio of this intensity to the amount of undertitration hecomes vanishingly small near the equiva-

I c IO

pump

I

I

The essential features of the discharge flow system are Figure I.. The electrode assembly was fabricated from ;O-mm i.d" Pyxex tube using W-Ni feedthroughs. Two identical pairs of electrodes (6 X 27 X 0.25 mm) were fabricated from stainless steel and spot weided to the Ni leads. The electrodes were seated loosely against the walls of the tube; the resulting spacing between the electrodes way ab'out 8 mm. The distance between the electrode pairs El and Ez was 5 ern. The electrode assembly was surrouirided by i t grourided chassis box. Polarizing voltages (up to 250 V) were provided by dry cells to exclude extraneous leakage paths between El and E2 or to ground. For routine measurements 25 V has been found to be satisfactory. kin currents were measured with a Keithley Model 4.17 picoammetes, having a sensitivity of 1 X A. Near-uv and visible radiation was :monitored by 1P21 pbotomulti.plier eirb~cwith Corning 7-39 and 3-66 filters, respectively. The :principal luminescence detected near the equivalence point of (1) was from the L" = 0 levels of the p s:ystern of N03J9-21 in the near-uv, and from the air aftergicvw in the v i s i ' r ~ k .The ~ limiting aperture to each photomultiplier was 6 mni locat)ed 8 em from the flow tube. :\Jz (Linde High Purity) was further purified by passage over C i l at 900 K, then through a glass-wool packed trap at 77 M, and then through P2OS.l9NO (Matheson) was passed through silica gel at 195 E( and P2O5 before being condensed at 77 K. The microwave discharge (2450 MHz, 100 W, %-wave cavity33) and the titration were carried out a t pressures of 60 to 1300 Pa. For a typical N2 flow rate (700 pmoljsec) m d pressure (800 Pa) the atom conS ~ Q W Din

I

I

,

I

I!

iA i

__

8 _ _ _ - -

consider here measurement of gas-phase electrical as a n addikional, independent indicator of the equivalence poirit of (.l). The presence of ionized species i s well kirown in the nitrogen aftergIow.9,14c,24-2~Ionization is substantially increased by partial titration with NO, indicating the occurrence of chemiionization reactions in this s y s t e ~ n . ~ 'The ~ - ~ primary :~ ionic reaction products are reported to be NO+ and free electrons, and the rate of ionization to be proportional to [NI3[O]. This concentration dependence has been interpreted in terms of a mechanism invcdying the participation of excited states of N2 and NO formed by atom r e c ~ m b i n a t i o n but , ~ ~ this mechanism has recently !been q u e s t i ~ n e d .If~ ~the reported third-order dependence of the ionization rate upon [N] in fact obtains, thttri ion formation would not serve as a useful equivalence point indicator for (1). We find, however, that the ion foolmation. rate exhibits a first-order dependence on [PSI mar the equivalence point, and is thus potentially suitable as a titration indicator. We describe an appara+,us for i.aeasurement of electrical conduction for this purpose a i d prment measurements demonstrating the applicability and samsitivity ofthis technique.

E,

h

Y

f

' &J

1

: I l = -

Block diagram of flow system and electrode assembly: B1, B2, ballast volumes ( 2 din3, 1 d m 3 , respectively); GW, glass wool plug; P J . P P I photomultipliers; E,, Ea, electrode pairs; V, flow regulating valves. Figure 1.

MW, microwave cavity;

centration was generally 0.1%, or iess depending on the amount of by-pass flow. Occasional experiments were performed with a trace of dry air added before the microwave discharge; the addition of air somewhat enhanced dissociation (up to twofold) but otherwise yielded results essentially similar to those without added air. Gas flows were determined from the pressure 'drop across a calibrated capillary,34 or directly from the pressure decrease of a reservoir of known volume. All experiments were conducted a t room temperature, 290-300 K. An air blower was directed on the flow tube just downstream from the microwave discharge to promote cooling of the heated gas. A glass wool plug removed excited species prior to the NO titration inlet.

Results and Interpretation ( a ) Mechanism of Saturation. In agreement with previous i n v e s t i g a t o r ~we ~ ~find . ~ ~ that electrical conduction in this system depends strongly upon the degree of titration. The conduction first increases as NO is added to the nitrogen afterglow and then decreases sharply as the equivalence point is approached. At low applied voltages f 5 0.5 V) the current depends linearly upon the applied voltage, but saturation is observed typically a t 3-6 '61, Figure 2, and the current increases only slightly (or occasionally is found to decrease somewhat) with increasing voltage as high as 250 V. Such saturat,ion might be due either to the formation of a plasma heath,^^,^^ at high ion concentration, or alternately to total extraction of ionic at low ion concent,ration. Hence it is important to determine the process responsible for the saturation observed in the present experiments. Examination of the current a t the downstream electrode pair E2, in the presence of voltage applied at the upstream electrodes El, should permit this determination. If extraction occurred at E,., then the measured. current a t Ez would be substantially reduced, unless the ionic concentration was rapidly restored by reactions of neutral species. However, the conduction of a shielded plasma would be minimally affected by prior application of an electric field. In Figure 3 are shown the results of siich an extraction study as a function of NO flow, for fixed N-atom flow, near the equivalence point of reaction 1. (This study was conducted without the ballast B2.) Upon application of polarization voltage to El, iz, the saturation current a t Ez, is substantially reduced. This reduction is a function of the NO flow relative to a fixed N-atom flow, but exceeds two orders of magnitude beyond the equivalence point, establishing that mobility limited e ~ t r a c t i o ntakes ~ ~ place The Journal of Physical Chemistry. VoI 78. fio. 1 1 . 1974

1122

S. E. Schwartz. and S. M. Butler

extraction of ionic species a t El is highly efficient, the present experiment demonstrates that the current i2'> measured a t E2 with E1 polarized, is due to ionization resulting from reactions of neutral species present in the flowing gas between E1 and Ez. It i s also established that 0 0 0 atomic oxygen in the absence of atomic nitrogen does not produce significant ionization under these conditions. It is further established that the current measured without polarization of E1 may be due in large part to residual ionic species in the flowing gas. The decrease in current at Ez was found to be independent of the direction of the polarizing voltage applied to El, ruling out the possibility that the apparent extraction was due to an interaction between the two pairs of electrodes. Occasionally, especially well before the equivalence point, there was a time lag of the order of several seconds before the extraction reached a steady level. In a I ! -:joL study of mass-spectrometric sampling of ions in active ni0 5 IO trogen Spokes and Evansz7 noted a similar time lag, which they attributed to surface charging effects. Despite VO!lS the time lag noted here, the response of the current at E2 Figure 2. Current-voltage plot in the flowing afterglow near the to changes in the NO Row near the equivalence point was equivalence point of reaction 1. The current was measured at E a , El not polarized. Ballast B:! was omitted; t h e contact time always quite rapid (better than 0.1 sec). ( b ) Debye Lengths. The possibility that the decrease in frorn the NO inie? otm EP was ca. 10 msec, [N2] = 800 Pa, F N ~ = ROO pmol/sec, atom concentration = 0.1%. The asymmetry apparent extraction a t undertitration is due to plasma is ascribed to stray fields, perhaps due to a voltage exerted by shielding a t the somewhat higher ion densities here, raththe uicoammeiter. er than to the presence of ionic species formed subsequent j to extraction, may be further examined by evaluating the Debye length. In order for a plasma to shield itself against an applied electric field it must develop a t each electrode a space-charge sheath, the thickness of which musk be several Debye lengths.a@ If the dimensions of the plasma do not greatly exceed the Debye length, then a sheath cannot develop and extraction will occur upon application of an electric field. Evaluation of the Debye length, h -7 ( ~ & T , / r w ~ re]~/~, quires knowledge of the electron temperature T,, and the number density of charged species, II. These quantities may be obtained in two ways. First, it is assumed that saturation is due to formation of a plasma sheath. then \ 3 the double-probe analysis35 may be applied. For the data D of Figure 2 , T, = 4500 K and n = 3 k IO4 m r 3 , and h = 3 cm (cf the electrode spacing 0.8 cmj. Hence the data of Figure 2 are inconsistent with the existence of a plasma, and it is inappropriate to apply plasma theory to interpret Figure 2 . Alternately, assuming extraction, it is possible to identify the measured current with the rate of' arrival of ion 3 pairs in the Rowing afterglow to the measuring electrodes. Thus n = i / e f >where e is the electronic charge and f is the volume flow rate. For the conditions of Figure 2 , n = N 0 flow/(Frnol/s) The electron temperature i s estimated as Figure 3. Ion and uv photomultiplier currents vs. YO flow: 0 , 1.5 x lo5 7000 li from the appropriate Townsernd factor3& j E / p 'i: ia, saturation current at Ea. El not polarized; 0, i:! , saturation current at Ea, E! po;larized (25 V ) ; 0 , uv photomultiplier cur1V rr-1 Pa-1 a t saturation); the resulting Debye length, 1.5 rent, arbitrary units; ballast BS was omitted, [ N a ] = 800 Pa, F N z cm, is consistent with the assumption of extraction, = 600 pmol/se?c. For currents of to A, the number densities calculated in this way ranged from IO3 to IO7 c m 3, and when El is polarized, and i s responsible for the observed the corresponding Debye lengths from 20 to 0.2 rm. Thus saturation. The magnitude of the observed saturation the apparent lack of extraction, measured at undertitravoltage i s con.sisten,t with that requisite for NO+ or anothtion, is due, as assumed above, to chemical. reactions of er ion of similar mobility to drift the interelectrode disneutral species, and not to plasma shielding. tance within the time spent between the electrodes, as (e) Effect of Contact Time. Before proceeding we point determined by the linear flow velocity. Before the equivaout the necessity, in gas-phase titrations, of allowing suffilence point the apparent extraction of ionic species by El cient contact time between the admixed gases for reaction is less efficient. The apparent transmittance of ionic to proceed to completion prior to measurement of a physispecies increases with increasing undertitration, so that cal property indicative of the equivalence point. in the exwell before the equivalence point there is but a slight deperiment shown in Figure 3 insufficient contact time crease in i2 upon application of polarizing voltage to El. (-10 msec) was provided, resulting in a trailing-off of Such an apparent lack of extraction has been noted pre~ ~ u ~ t both the ion current &', and the iiv ~ ~ ~ ~ t ocurviouslyZfifor active nitrogen. Since it is established that 7 f

N

The Journal o l f h y s i c a l Chemistry. Vol. 78. No. 11. 1974

Equivalence Poinl Determinarlon by Electrical Conduction

1123

f'

s S

\ z)

\ - M

.-

P

.Oil0.6

MB f l w / ( p o l / s )

ure 4. Ion and uv photomultiplier currents near the equivalence point of reaction 1 . Et wa,s polarized; with ballast B2 the contact time was 0.5 sec: l@, i2 , ion saturation current; 0 , uv photomultiplier current; I N 2 1 = 1000 Pa; F N =~ 800 Mmol/sec.

rent, I,, (reaction 31, beyond a n apparent equivalence point obtained by linear extrapolation of Zuv measured before the equivaience point. With an initial atom concencn~-~ the , half-life of reaction 1 is initration of 4 x .~~ for equivalent flows of the two tially 0.1 r n ~ e c However, reagents the rate of reaction decreases as the reaction proceeds in proportion to the remaining concentration of reagents. Thus, for example, 10 msec is required for 99% completion, even if perfect mixing of the reagents is assumed; the traiYing-off in Figure 3 may thus reasonably be attributed to the finite rate of reaction 1. For this reason we have inserted the ballast Bz (Figure 1) between the NO inlet anti the point of monitoring the reaction. This bulb provides an average contact time of 500 msec under typical conditions. ( d ) Dependerzce of Saturation Current on NO Flow. A series of studies was conducted to determine the dependence of the saturation current iz' (measured with the ballast Bz, and subsequent to extraction a t El) upon the flow rate for fixed N atom flow rates. In these studies the NO flow was established somewhat in excess of the N atom flow arid allowed to decrease as the reservoir pressure decreased.:* T h ~ tmeasured NO flows were fit as a function of time to allow the ion and uv photomultiplier currents be ,jeternlined as a fuIlction of the NO flow rate, FNO (Figure 4). Both currents exhibit a sharp onset that occurs a t the same NO flow within an uncertainty of better than 0.170. This agreement was observed in all instances except a t the lowest pressures (