INTERACTION OF NITROUSACIDWITH HYDROGEN PEROXIDE
Dec. 20, 1054
The process controlling the rate of formatiotl of phosphorus nitride from phosphorus and nitrogen a t temperatures above 1800°, therefore, is viewed as the catalyzed dissociation of diatomic phosphorus into atoms. The process probably involves adsorption of diatomic phosphorus on the tungsten, Ielease of phosphorus atoms from the surface and reaction of the phosphorus atoms with nitrogen in the gas phase. The ratio N : P in the phosphorus nitride products was always greater than unity. This occurrence could result from the presence of phosphorus atoms in accordance with the mechanism postulated below. Phosphorus atoms-produced a t the tungsten surface-react with nitrogen molecules to form the complex PNN, which then reacts with another phosphorus atom. The resultant groupings may combine to yield the polymer (PN)n. Or they may continue to grow through alternate collisions with nitrogen molecules and phosphorus atoms until the reaction ends a t an N : P ratio in the range from unity, as in (PN)., to 5 : 3 , as in the polymer (P3Ns)%,which has the highest nitrogen content of the known phosphorus nitrides. Other mechanisms were considered : (a) the atom chain reaction
6243
of P N indicate that the energy level of mechanism
(a) is nearly twice as high as the experimental activation energy. The chain reaction might proceed, however, in the vicinity of an electric arc. The magnitude of the experimental activation energy, 59 kcal., is inconsistent also with mechanism (b), because the formation of PZmolecules requires only 26.8 kcal. and the estimated energy level of a PzNz complex is considerably below that of phosphorus atoms. The reaction velocities found experimentally were compared with those calculated according to the absolute rate equationz7 which applies to the postulated mechanism
The number of active sites per square centimeter of catalyst surface, cs, is a function of the composition of the catalyst and the degree of covering by the adsorbate. From the adsorption isothermz7applicable to the mechanism suggested here, the degree of covering is estimated to be 1 to 10%. An assumption that the phosphorus-tungsten complexes are the active sites leads to absolute reaction rates five to ten times as great as the experimental values-a fairly good agreement in view of the uncertainties as t o the surface area of the catalyst P N2 +P N N and the efficiency of condensation of the nitride. N Pz +P N P Acknowledgment.-Henry Eyring, Consultant leading into a polymerization reaction, and (b) t o TVA, gave helpful advice concerning the study. reaction of the diatomic molecules Mary Griffin Rountree and A. J. Smith assisted with the experimental measurements. J. A. BrabP2 N2 +PzNzZ +2PN also leading into a polymerization reaction. Cal- son, 0. W. Edwards and Inez Jenkins Murphy culations based upon the dissociation energies of NP dealt with the problems of chemical analysis. and P N and the probable energy of polymerizationz6 Thad D. Farr made useful suggestions about the manuscript.
+ +
+ +
+
(26) National Bureau of Standards, "Selected Values of Chemical Thermodynamic Properties," Series I, Vol. I, Table 19-10, Washington, D. C.. December 31, 1947, and Marc0 31, 1949.
[CONTRIBUTION FROM THE
(27) Reference 18, p. 358.
WILSONDAM,ALABAMA
KENTCHEMICAL LABORATORY, UNIVERSITY
OF CHICAGO]
Interaction of Nitrous Acid with Hydrogen Peroxide and with Water BY MICHAELAN BAR^
AND
HENRY TAUBE
RECEIVED JUNE 24, 1954 The rate law for the exchange of oxygen between nitrite and water in the p H range 4-6 is k(HC)*(N02-); a t p = 1.00 and 2 j 0 , k = 2.6 X 108 1.2mole+ min.-l. The rate is unaffected by phosphate buffer a t low concentration. The form of the rate law, the inhibition by HZOZof the exchange, and the kinetics of the reaction of NOZ- and HZOZare explained by the formation from the activated complex of an intermediate (possibly NO+), and competition of HzO and HzOz for reaction with the intermediate. At low nitrite concentrations, two oxygen atoms are transferred from peroxide for each mole nitrate
**
formed, corresponding t o decomposition of an intermediate pernitrate (ONOOH) by internal rearrangement. At high nitrite and low peroxide one atom of peroxide oxygen appears in each nitrate formed, a result which corresponds to the proc-
+
+
ess: OiYb8H Not--+ OiYO**ONOz- f H f . When nitrite and peroxide are high, peroxide oxygen appears in the nitrite (but nitrite oxygen does not appear in the peroxide). The exchange results are applied in testing mechanisms which bave been proposed for the reaction of amines and nitrites.
The reaction Nos-
+ Hz02 = 3'01- + Hz0
takes Place very rapidly in acid solution, but the rate decreases markedly as the acidity is reduced. The rate law for the reaction has been shown to be2 (1) UNESCO Fellow on leave from Weizmann Institute, Rehovoth, Israel. ,(2) E. Halfpenny and P. L. Robinson, .7. Chem. Soc., 928 (1962).
-d(peroxide)/dt = k(H+) ("02) (H202) with k = 8.3 X 1031.2mole-2rnin.-' a t 19". It also has been shown3 that in acid solution pernitrite (presumably ONOOH) is formed in equivalent amounts. This was done by making the solution containing NO^-, H ~ O and~ acid alkaline after mixing. I n Some experiments as much as 0.7 of the nitrite (3)
K.Gleu and R. Hubold, 2. anorg.
Chcm., 22S, 306 (1935).
MICHAEL ANBARmr) HENRYTAUBE
6244
appears as pernitrite. The pernitrite persists in alkaline solution, but decomposes rapidly in acid to, yield NOS-. It seems reasonable to suppose that pernitrite is an intermediate for reaction (1) at moderate acidities, even though not enough is present a t the steady state under these conditions to make its presence easily recognizable. We have been concerned mainly with experiments on the system using OI8 as a tracer. The slow exchange of HzOz and Nos- with water makes it possible to obtain meaningful results in such a study. The tracer results are especially searching in the present system, because the reactant NOaand water can also be used isotopically distinct in composition. They are given added interest by comparison with those obtained in the reaction of sulfite and peroxide,4 which showed that each sulfate formed contains two oxygen atoms derived from the peroxide. Experimental Definitions and Conditions.-Time is expressed in minutes, and concentrations in rnoles/liter. N represents the mole fraction of 0 ' 8 in a particular species. Subscripts represent the time of reaction. E represents the enrichment ratio, ;.e., N in a sample compared t o N for a sample of normal isotopic composition. n represents the number of oxygen atoms derived from peroxide in each nitrate ion formed in the reaction. The temperature of reaction in all experiments was 25 rt 0.2". Materials.-Chemicals were A.R. quality. Water, used as solvent, was purified by redistillation from alkaline permanganate. HzSO~in solution was obtained by the r a w tion of K&08 and sulfuric acid. Solutions used soon d t e r preparation contained less than 5% of the oxidizing a : p t as HaOs. Procedures.-The oxygen isotope composition of the 16trite was determined using the reaction with azide ion HN3 HSOz NzO f !S* f Hz0 and analyzing the N20 in the mas5 spectrometer. Other work has shown exchange of NzO with solvent t o be very slow.6 The reaction medium Tyas a phosphate buffer a t pH 2 . 5 , and the reaction was conducted, adding the nitrite to the azide contained in the buffer. When p H is much lower, the exchange of nitrite with solvent during the reaction becomes too great; a t higher pH, the reaction is too slow. Attempts to analyse by acidifying mixtures of Naand NOz- contained in alkali revealed a remarkable result: the exchange of NOZ- and water in alkali is ordinarily very slow, but with azide present it takes place rather rapidly. The oxygen isotope composition of nitrate was determined on NzO formed b y the reaction of NH4+ with NOa-. TO apply this method t o nitrate formed in the reaction mixture, all deleterious substances were first removed. Excess nitrite was destroyed by hydrazine, sulfate and phosphate were removed as the lead compounds, using PbClz as t h e precipitant. Solid iYH4C1 was then added, the solution adjusted with sodium hydroxide to a faintly acidic condition, and water was removed by evaporation a t room temperature. The reaction mixture, containing mainly Na+, NOJ-, NHa+and C1- was dried thoroughly, transferred t o a platinuni crucible, then heated a t ra. 300" to liberate NtO. This gas was first condensed over solid NaOH t o remove a n y acidic substances formed as side products, then over concentrated sulfuric acid, and collected for mas3 spectrometric analysis. Isotopic analyses of the hydrogen peroxide were made, oxidizing i t Kith Ce(1V) and examining the isotopic ratio in the oxygen. The isotopic enrichment of the stock enriched water was determined by comparing isotopic ratios on COSequilibrated with 0 ' 8 enriched and with ordinary water. The enrich-
+
16
VOL
ment ratio of water in any solution was estimated, taking account of the dilution by ordinary water, and any oxygen exchangeable species. The PH values of solutions were determined with a Beckman meter, model G. Calculations.-The experiments on the exchange of SOSand solvent required measurement of the isotopic conxpoition of i S 0 2 - as a function of time. The logarithms ofi (Nt N,)/(No - N,) were plotted against time, a n d the. specific rate determined from this plot is related to the r.rfe, R , of the reaction carrying the exchange by the equation
-
where R is expressed in the units moles of 0 per liter per minute. The experiments on oxygen atom transfer required comparison of the isotopic composition of the nitrate formed with that of the nitrite and peroxide undergoing reaction. The hydrogen peroxide used was of normal isotopic composition, but the solvent, and the nitrite in isotopic equilibrium with i t were enriched in 0l8. To correct as far as possible for fractionation effects (these arise as a result of equilibrium isotope discrimination in the pair R'Oz--HtO, the fractionation of oxygen in the reactions to produce NzO, and as a result of side reactions during liberation of NzO), the oxygen isotopic composition of nitrite in equilibrium with solvent of normal sotopic composition and that of nitrate formed from it by reaction with hydrogen peroxide of normal composition were compared, and the ratio found (1.04) was applied to a l l the data. The greatest source of error in the rate data appears to be in the determination of (H+); in addition to the difficulties in principle with using pH as an index of this quantity, is the imprecision in the measurements. This imprecision is magnified in the values of R , since they depend 0 1 1 th?. .square of (H +) .
Experimental Results The results on the exchange of NOa- and solvent are presented in Table I. Each value of R entered is calculated from a slope determined by 4 or more points. The lines through the points were straight, indicating, as was shown directly, that there was no appreciable change in PH, and that net decomposition was only slight. TABLE I THEEXCHANGE OF OXYGEN BETWEEN NITRITEA N D WATER Phosphate used as buffer except in 1.09, 1.61 and 1.62. Buf(KOA-) fer 0.11 1.01 0 . 2 5 .I1 1.02 .25 11 1.03 .25
Exp
1. 0 4 I .05 1 06
1.07 1.08 1 09 1 10
111 1.12
1.31 1.32
1 41 1.42 1.43 1.44 1.51 1.52 1.53 1.54 1.55 1.61 1 ti:!
.2.5 25
.25 .25 ,25 ,25 , 2.5 .23
.I1 .l5 .10 .IO .IO
Other
,......... ...... ... ...,...... ,.........
..........
NaClO,, 0 . 2 2
,25
10 , I1 .I1
.4.5
.91
,114
.91
.30
. .
. ...
NaCIOI, 0 34
,10
09
.I0
.09
... ... ...
... .,.
8.00 . 4 4 0.30 .41 6 . 3 0 .45 6 . 0 0 .4A 5 . 9 5 . 3 7 5 70 .35 5.10 .35 5 . 0 0 .32 5 . 0 0 .57 .35 .32
1.36 1.36 .5.5
0.0.5
025 .IO .10
fiH
0.69
.IO"
.50 .20 .20 .00.5 .Ol ,05 .IO .I0
p
.. ,. .. .. . NaBr, 0 , 2 NaClO4. 0 . 2 NaClO4, 0 . 9 9 NaClOd, .99 NaCIO4, . 9 5 SaCIO1, . 9 0
.55 , AG
.50
4.67 4.40
0.0003 . OX8
,031
4.53 4.43 4.35 4.30
?? .i2
3 75 5 40
o in
a x 1fi-a
100
c
4.0
5 5.6 4 .9 6.0 2.7
4.9
2.0
3.8
1.5 2 0 2 0
.I? IS
.50
6.3 21 39
5 . 0 0 31 9.9 5.00 2.1 5.47 4.6 5.30 0,s.; 5.70 0.43 5.67
1.00 1.00 1 00 1.00
........., C.Ii~YH:, 0 . 1 C r l H s S H 2 ,0 . 4
5.00
R X 10s
1 .no
3.3 26 62 RO
(1.1;
7.1 8.8 3G :i .7 4.4
?.E 2.4 2 . (i
2 8 2.3 1.8
(I .i
l i
I,
I ,
;I
b
6245
INTERACTION OF NITROUS ACIDWITH HYDROGEN PEROXIDE
Dec. 20, 1954
The values of specific rate entered in the last column of the table were calculated from the rate law R = k(H")'(NOz-)
sponds to the band maximum for NOz- was used to follow the reaction. This wave length is much more strongly absorbed by the pernitrite' than by nitrite. There was no evidence for the growth of pernitrite, indicating that its stationary concentration is very slow. The system is well behaved kinetically, and in buffered solutions, the reaction is strictly first order in NOz-, to a t least 90% decomposition, which was about the limit for meaningful data. The specific rates tabulated are defined by the rate law
In this rate law, (H+) = antilog (-$H), and NOz- represents the concentration of nitrite ion, calculated from the stoichiometric nitrite, allowing for the association to form "02. The value of K d i s s . 6 used for this calculation was 5.0 X a t all ionic strengths. In the most extreme case, a t the lowest pH studied, this correction amounts to only 16%. The first fourteen experiments described were performed before a new method of making isotopic An attempt was also made to obtain data for s y s analyses on NO2- had been worked out. The values of R entered for these are not as precise as they tems a t low acid, not containing buffer. The curves are for the remainder of the experiments. in a plot of log (extinction - extinction) vs. t were Some experimental results on the exchange of found to be concave down. This feature is readily oxygen between NOz- and HzO in the presence of accounted for in the following way. When (NOz-), Hz02 are recorded in Table 11. I n these experi- >> (HN02) >> (H+), for the major portion of t h e ments, nitrite and peroxide were of normal isotopic reaction, ("02) remains constant as NOz- disapcomposition and the solvent was enriched in H2OIs. pears. Thus it follows that (H+) = K(HN02)/ (NOz-) varies inversely as ( N ~ z - ) and , dhe rate TABLE I1 law k'(H+)z(N02-) assumes the form -st./ R k ( N o r - ) Buffer (Hz02) EXP. CI X 108 X 10' $H (Not-) with const = kK2(HN02)2. The d&ufor 1 . 0 0.45 4.75 2.01 0.2
