Title Emission spectrum of the flame of bromine burning inhydrogen and mechanism of the reaction

Author(s) 北川, 徹三

Citation 物理化學の進歩 (1938), 12(5): 135-147

Issue Date 1938-10-31

URL http://hdl.handle.net/2433/46156

Right

Type Departmental Bulletin Paper

Textversion publisher

Kyoto University

9fiT~lt~olit€1 Vol. 12n No.

EMISSION SPECTRUM OF THE FLAME OF BROMINE

BURNING IN HYDROGEN AND MECHANISM

OF THE REACTION*.

By Tsrsvzo K~rncnwn.

In the preceding paper" it was reported that in the emission spectrum of the flame of chlorine burning in hydrogen some emission bands were found, which

were ascertained to belong [o the excited chlorine molecules, and the mechanism of the combustion reaction between chlorine and hydrogen was discussed from these

results.

In the combustion of bromine in hydrogen a number of emission bands belonging

to the excited bromine molecules were found. In the present paper the mechanism

of the reaction between bromine and hydrogen will be discussed front this fact

[I] The Emission Spectrum of the Flame of Bromine in Hydrogen.

(i) Apparatus

The apparatus used is shown in Fig. t. Purified hydrogen was introduced through a flow-meter and the side tube

0 ~ e into the quartz vessel a, 3cm. in y di

ameter and iocm. in length, and the

~1 1 ~.. ,~

Fig. a. The Dtain Par[ of the Appara

the temlxrature of the rescn'oir of

boiling point of bromine (53.y°C.) b

(BS, No. 38) coiled round it. Th method to light the flame of bromin

No grease was used on the

ground part B was joined with a The same glass spectrograph

• '1'Itis paper is the Lnglult irar~slati 10, r (r936).

r) 1'. Kilagawa, Hrv. Pbys. Chem, Jn

diameter and iocm. in length, and the

."'~..~-- excessive hydrogen and [he hydrogen

bromide formed by combustion were

expelled through the other side tube d.

Bromine vapour tvas let flow out Iw

through Band /~ at a constant velocity,

liquid bromine 6 being kept constant above the

y means of the electric current on nicltrome wire

e tube B-Is was specially made of quartz. The

e was tUe same as that of the preceding report".

passage of the bromine vapour and only the

small quantity of viscous phosphoric acid.

was used as in the preceding cxperimento. The

un .°f the same article pohl~hed in Rei. P/rys. Cluua. Japan,

prru, 8, 71 (F934): II, 6r (a937)•

5 (1938)

I

i 1

c

r

9fi~~lt~mit€~ Vol. 12n No. 5 (1938)

136 T. KITAGAwA Vol. Xtf

dispersion of the spectrograph was t a P,/mm at 450o A and q5 fl~mm at .i 590o A.

The Ilford Hypersensitive Panchromatic Plates were also used.

(ii) Temperature of the flame

The temperature of the flame was measured by means of a thermoelement

with plafinum-platinum rhodium alloy, which indicated the temperatures of about

430°-..5oo°C. There was some reason to believe that the temperature of the flame was somewhat higher than this. It was proved, however, that the temperature

.of the flame of bromine. was relatively low compared with the, flame of chlorine which had been about tooo°C.

(iii)

The

is shown

Emission

flame was

in Plate

spectrum of the flame

visibly of orange colour. The emission spectrum of the flame

I, the time of exposure being about t hour. In Plate I about

Hand-Heads in the Emission

Table L

Spectrum of the Rrominc-ifydrngen •Elame~

No.Wave ]ength

1, ~

Wave mm~hei

v, cm-t

Int No.Wave length

a, A

Navc numbee

Y, cm-~Int.

I

z

3

4

5

6'

7

8

9

Io

Ix

t2

t3

[q

x5

xb

t7

t8

xg

zo

2I

zz

6875

680[

674q

67z9

6675

6667

6605

6593

6545

6532

6474

64x7

6406

6360

63gz

63[3

6ago

6z6q

6zgo

bzw

6159

6165

t45gt

x4700

Ig5z3

x4557

x4975

x4995

x5x35

x5x63

I5z69

x5305

x5443

x5579

x5605

I57I9

x5763

x5336

x5593

[596x

x6ozt

I6o7z

I6t55

t6zog

0

t

3

3

5

5

7

7

S

S

xo

to

6

9

7

S

9

6

8

3

7

5

z3

z4

z5

z6

z7

z8

z9

30

3x

3z

33

34

35

36

37

3S

39

40

4x

4z

43

6[qz

6tzo

6097

6o7z

6056

6ozq

59Sz

5957

594x

5gto

5567

583x

550z

5794

576t

5753

57z4

5692

5660

5629

56oa

x6z76

x6334

16395

t6g66

t65o8

16596

I67tz

t678t

x68z7

16916

x7039

[7146

[7z3o

17z54

17354

x7377

17467

17563

17663

x 776z

t7S5z

6

4

5

4

z

4

4

3

3

5

5

4

t

3

z

x

z

I

1

0

9fiT~1t~~i~1 Vol. 12n No.

No. b' L\SISSION SPEC1'RU\f 4lP T1IE FLAML•' UP DRO1.ffNE 137

- - ~ '

Plate L

TlieEmissia0 Spectrum of [Ite Flame (Rel., Fe-arc spectrum}

43 emission bands are observed in the range A.I 68So-.-Sooo A. The fine structure of each band is obscure, but the band-head is distinctly observed, and all the

bands shade to the red. The band-heads of these emission bands were measured,.

being referred to the Fe-arc spectrum. In Table I are given the wave-length d,

the wave number in vacuum v and the intensity roughly estimated.

(iv) The carrier of the emiaeion banda

Front the fact that the emission band system in Plate I is relatively of simple

structure it is probable that the carrier of the spectrum is a diatomic molecule.

The diatomic molectiles present in the flame are hydrogen, hydrogen bromide and

bromine. The band spectrum belonging to hydrogen is known as many lined

spectrum which exhibits innumerable emission lines and is not of an ordinary band

form. Hydrogen bromide has a continuous spectrum as far as we know. Weizel,

~VolEf and Binkelepl found in the discharge tube of hydrogen bromide a wavy

continuous spectrum in the ultraviolet region, but no emission spectrum corres-

ponding to this could be found in the flame of bromine and hydrogen. It is

evident, therefore, that the emission bands in the flame do not belong to these

hvo kinds of molecules. Accordingly, it is considered that the carrier of the bands

will be the bromine molecule.

Many researches on the absorption spectrum of bromine vapour have been

made. Let us compare the absorption band of bromine with the emission band of

the flame in question. The tvavc length. of the absorption band of bromine was

measured by Kuhn"I, Nal.amura" and ]ately 13rown't. In Table II are given the

wave number by Bro++'n and that of the emission band by the author in the second

and the third column resp. Comparing these values it is found that both wave

numbers coincide well with each other within the experimental error. This leads

to the conclusion that the emission spectrum of the Game is ascribed to the bromine

z) \C. \CCizcl, 1[. R'. l\~olff u. IL F.. liinkcle} 7.. j./rwik. Ckrur. (lt), ]0, 459 0930)• ;) IL Kukn, %. F/iyrik, 38. 77 (+T-6)•

q) U. Nakamura, Jlrn:. Colt Sr. d'yd[u Guf. U+eiv. (.1), 9, 335 ft926} g) tY. U. Brown, Pk~~r. Rr~., 38, n79 R9Y); 38, 777 (+93~}

5 (1938)

1

1

I

9fiI~1t~~i11~ Vol. 12n No. 5 (1938)

138 T. KITAGA\VA Vol. XII

l.9mpa~ISOn of the Absorption Rends

Table II.

of Rmmine with the Emission Rinds of the Flame.

No.

Absorption bands n( Lmmine

(by L. Brown) cm-t

Vibralional

quantum number (d. r~~)

L•'musion Lsnds oC the flame

(hy the author) pn-t

Calrulated valnc

em-t

DilFerenee

ea, ~

11-6

15

to

3-to

7b

7

7

78

7zo

-i

08

5886z

46

Io

3i80

374

-3-

,7zzfi

-3_g

60

10I76

-x-3

[455z•4

14694-0

1471461483z.9tg86o.114967.815aoz.t1514o.G15 t7o.z15275.515312.215450.715585.6[562q.G157t6.S1576;:115544315898.015969.1t6oz9.216077.716156.7[6212.61(x81.516343.816400.51647[316516.61 C.5~.116715.116788.3t 6831.216913.117o3z.i17[gS.zt 7x32.317260.417351.317368.817473.017467•~l17573.117579.617669.117761.017848.6

14541

147ao

[¢7ao

1432314857

1497814995

15135

15163

15269

155

1544315579

Ct5605]15719157631583615893159611692116o7z16155t 62o916276163;416398164661650816596167 i2t678t[6827169161703917146t7x3o172541735417377174671746717563

[t7563~1766317762t78~z

15169.2 (6,5)

15450.0 (8,5)

15585.5 (9.5)t56zq.6 (7A)15714.7 (10,5)157626 (Sq)15845.1 ([1,5)158g8.S (9,4)15967.2 (12,5)t6oz9.q RO,a)16078.4 ($3)16t56.4 (11,4)t6z1q.0 (9,3)16278.9 (iz,4)[6343.4 (10.3)16go1.q (13.4)16472.2 (t1,3)16517.5 ([4.4)16594-0 (tz,3)167t8.7 (13,3)t673q.1 (11,2)[68;[.5 (14.3)t69IZ.9 (1 z, z)17033.7 R3."-)17150.0 (14,2)

17259.8 (15,2)173533 ([3,1)t7368.q (16,2)17472.1 (17,2)I7467.t (14,1)17572.0 (18,2)17579.4 (15,1)17667.6 Q9,z)17759.9 (~,z)17845•S (z1,z)

I

2txrs

34

56

7x

9Io

l1

121.31q15t617ISt9zoxtzzz3z4a5z627z8z9

;o31.3z33343536373839139rs401qo.1414z43

9fii~lt~mitt~ Vol. 12n No. 5 (1938)

No. 5 RMi5510N SFECTRIJNi OF THIi FLAME OF BROMINE. 1:19

molecule. This conclusion

band of the flame and the

is also supported by the (act that both the emission

absorption band of bromine shade to the red region.

(v) On the emission band of bromine

Fety studies on the emission band of bromine have been done as compared with those on its absoption band, and any emission band system of bromine which

coincides well with the absorption has not yet been found`) in the discharge tube.

According to Urey and Bates'', a faint band spectrum in the range behveen .i.I g25o and gG75 A was found in the flame of oxygen which contained hromine

burning in hydrogen, and the spectrum u'as considered to belong to a compound

of oxygen and halogen. In the present experiment, however, any such emission

band tvas not observed in the wave length range.

According to R. S. Mullikeno, the emission band of bromine in the flame

(Plate 1) shoud he due to the electron transition from the Ou* state to the t£R state, for the normal state of a bromine molecule is '_`'k and its excited state Ou*. Accordingly, the excited bromine molecules in the Ou* state must be generated

in the flame.

Before we consider how the bromine molecules arc excited by the chemical reaction in the flame, let us calculate from aband-head analysis the vibrational

energy which is possessed by the excited bromine molecule in the Ou* state.

[ll]

(i} Band-hend

Band•Head Analysis and U(r}Curves of

analysis of the emission band

Br:.

The band-heads in the emission spectrum (Table I) were analysed and the

results obtained are given in Table I11, where the unit is given in cm ' and the

intensity is indicated in parentheses. J and ?r' in Table III are the \~brational

quantum numbers of the excited and the normal states of the bromine molecule respectively, and these values have been determined in the isotopic displacement

of the absorption bands by Brown.

From the band-head analysis it is seen that the emission bands of bromine in the flame have been emitted by the transition from the vibrational states a/=

6)

7) S)

Sa)

Y. Uchida and V. Orn, fnp• J. Phrs., 5. 59 ([9zS). IL C. Umy and J. R. hate., Phrs. A'ri., 34, [54t P9z9)• R. 5. ]Cullikcn, Rr~. AIa/. Phys., 4, t (t93z). \\'. !1L Vaidya has lately olbened in the flame of ethyl hmmidc the envision hand; of hromine ad.ieh wincide with the pceunt author's (Pror. Jn~. Atad Sn: [A), 7, 3z1 ([93S)}

9fi7~(t~mi~~ Vol. 12n No. 5 (1938)

740 T. KITAGA\\'A

Ta61c III.

Band Analysis n( the F.missinn T'.an~ls in the I•lmue, []k_,

(J and r~rr arc the bihratinnal cu:odum numbers n( Ibe and the normal states respectively.)

excited

VnI. XT1

r a 7 4 5 fi

5

G

7

8

9

Ia

II

I2

I3

iq

IS

16

I7

IS

[9

zo

z[

n

Ou•

x7z3o(x)

x7354(2)

x7467(2)

x7563(U

x67St(;)

x69tG(5)

x7°39(5)

x7[46(4)

xi 254(3)

=7377 U ) x7467(2)

t7563(x)

x7663(:)

t 776a(t)

x7852(0)

x6°72(3)

x62°9(5)

t63;W(4)

x6465(4)

xrs96(a)

x67+z(q)

[6527 (3)

x56°5(6) +5763(7)

x5593(9)

[6ozt(S)

xb+55(7)

thz7r,(6)

xh393(5)

[h5o8(z)

+5+6.;(7)

x53a5(S) x5443([01

+5579([0)

x57x9(9)

t5S3G(S)

x596x (6)

xa7~h) xass7 C;)

+4995(5) x Sx 36(7)

x5269(5)

x4s4+(~)

I47oo(+) xg8z3(3)

t497S(5)

~r~+

eri'

aoow

imoo

U

'ik .'SsO_

E.i

0 ~4 L6

Bfi

A.'Pb

[t

r

20 14

Fig.

L 31 16

2. f1~r}Curses of the

40

Rr_

49 9!

molemlc.

10 San

9fiT~1t~0it€~ Vol. 12n No. 5 (1938)

No. ! F.DtISSTON SPECTRCIM OP TI-IE FLAME OF BROMINE, 141

5~.zt in the excited state to J'=t-..y in the normal state. Thus, the vibmtional

quantum numbers s/ and v" of the emission bands could be determined as follows

The tivave number v of the bromine

bands can be expressed by the followinb

formula given by Brown:

v=t583t.z+(163.8tJ - t.59v''"-0 .0087"?-(3=z.71 u'

The calculated wave numbers being st[bsti-

tuted the values of v' and z/' (1) in formula

(z) are given in the fourth column of Table II, which coincide well with the wave num-

bers of emission bands of the flame in the

third column.

In Table IV the bands observed by IIrown in the absorption spectrum of bromine

vapour ate represented by the sign O and

those by the author in the emission spectrum

of the Elame by the sign + for simplicity.

{iI) D(r)<orves of Br:

In order to make clear the mechanism of the emission of the bromine bands let us

draw the U(r)-curves of the bromine mole-

cule. The formula to express the potential

energy U in a diatomic molecule has been

given by Morse'1 thus :

rt= ./gx=exetutµ~k em '. (3)

The vibrational energy G(v) of a di-atomic molecule whose vibrational quantum

number is v is expressed by

z \ 2

9) 1V. Jevons, •' Refa# wr Bmd-Speetrn of Dinfai

Table IV.

Simplified Expression of [he Band Malysis

oT Bromine in the Transition Ou+~~£g+.

O Observed in the absorption spectmm 6y G. Brown.

t Observed in the emission spectrum of

the Bre-H~-flame by the author.

o l x l z 3 4I5 6 7

0

x

z

3

4 O

i O

6 O O O

7 O O t

S O

9 O t

xo O

v

xz

x3 m

xq

x5

xfi O O

x7 O O

:8 O O

c9 O O

zo O O

z~ O O

zz O

z3 O O

za O O

48 O

~~~+ 2 ~-x`W`\y+ 2 /E-I-y`tu`(.~} 2 /a ...gym ~, ~4)

wi Bmd-Speetrn of Dinfauit dlaleculec;' I93z.

9fi1~(t~mii>1 Vol. 12n No.

142 ~ T. X(TAGAIVA V01. XII

where yrra, is so small that the third term in the' right side can be neglected when

v is not so large"~. F.,, D„ rur, r~rur, r, and a in these formulae are given in

Table V.

Table V.

Mnleculnr Cnnstnnts n( the Excised and the Nnrmn7 Stntes n( the lirnminc h(nleculd^1.

Cnnstnnts I E. I D I De me I xrou~ re I n

5 (1938)

Units cm-t eels

Ou•-atnte

t£g}-stele

ISgro.S ~ o.gbz 1.961

kcal

Io.65

q;.m

cm-t

3gz7.3

10057.5

cm-1 cm-t

165.39

3z3.86

1.59

t.15

Io-sae

zcs

1~

totem ~

r•947

t.655

The molecular constants in Table V have been adopted from Jevons' ReporP't. The U(r)-curves of the bromine molecule obtained by equation (3) is shown in

Pig. 2, where the potential energy U (cm ') is taken as the ordinate and the

intenutclear distance r (cm) as the abscissa. The upper curve indicates the cscitcd

state and the lower one the normal state. The calculated vibrational levels arc shown by horizontal lines in the figure. The arrows downwards indicate the

processes of khe transition for the emission bands of bromine in the f3ame. The wave lengths of the emission bands starting from the vibrational states

such as W~5 are, from Franck-London's principle, so long that they go beyond

the sensitive range of the photographic plates used, and from this reason they

have not been observed in practice. From Table III it is apparent that the inten-sities of the emission bands belonging to d>[~ are very small. It follows, there-

fore, that the excited bromine molecules at the Ou* state in the Flame are, in the

main, in such vibrational states a J=[5~o.

[III] Chemical Reaction Mechanism.

(i) Mechanism of the emission bands of the flame

It has been clarified in the preceding section that in the reaction system of

the combustion of bromine in hydrogen there must be formed the bromine mole-

cules which have been excited to the Ou* state. These molecules will be written as Bro• hereafter. Let us consider how Bn* is formed in the chemical reaction

between bromine and hydrogen from the standpoint of the chemical reaction kinetics.

(t) Hoty much excitation energy is required in the formatimt of the Br_`? The inner energy of a diatomic molecule contains, in general, electronic, vibra-

to) In this TaLlc D=P~ G(n), inhere C(n) i. the vibrational energy nt the nhsnhtle zem tcmpnra~nm. II) \V. Jevons, "/infnrl ar Rmrd-Sfedm of Dinlnntic JJnk•rrrlrti' r9.i~. P• z~

9fi1~1t~mi~1 Vol. 12n No. 5 (1938)

No. 5 EMISSION SI'EC'1'RUM OF TI-IE FLAME OF EROMINF. 143

tional and rotational energies, of which the quantum of the rotational energy isso

small as to be neglected here.

According to Jevons'°, the electronic energy E,„r required to excite the bromine molecule from the r`'R state to the Ou* state is 45,ooocal. (v'0~D1=L533[crn '). Let

F..f and Er/' represent the vibrational energies of the molecule in the excited and the normal states respectively, then the required excitatiat energy [~ is expressed

thus

The vibrational energies Ev/ and F_s/r can be calculated by formula. (4), using the

molecular constants given in Table V"'.

(2) The origin of the excitation energy YV-should 6e brought forth from the chemical reaction heat Q. If so, the amount of the reaction heat has to be enough

for the :excitation energy. It has been admitted by many experimental results that the reaction

BrQ+ H_=?HBr + 24, 3oocal. (6)

proceeds according to the following reaction mechanism"': BrY=2Br-¢g, zoocal., (7)

Br+H._=HBr+H-t6,2oocal., (3)

and H+Br,=HBr+}3r+4o,5oocal. (q)

The bromine atoms formed in the primary reaction (7) bring forth successively the

secondary reactions(3) and (q).

In the reaction system at higher temperatures such as the flame, it can be

considered that the alternate repetition of the reactions (3) and (g), namely chain

reactions, may take place. In the chain reactions the reaction (q) is only an exo-

thermic one. It follows, therefore, that the excitation energy FV must have originated from the reaction heat . produced in (q). Then, is the reaction heat Q=

4o,5oocal. of reactiat (g) really sufficient for hV? a) 1f F_r+=o in equation (5), then

where id~'(o) represents the energy required to excite Bre to the lowest vibra-

tional state of Br._*`. It the reaction heat Q is considered to be not smaller than FV(o),

Iz) F.v-.frc•C(c) erg., and hetw-een the units of energq there holds the (allowing relation: x volt =Sioficm t=z3,o55cnt=i.53X io-~-ergJmelecule.

73) Ih. F. TlonlmelTer u. I'. Etarteck, „Cnnrdln~~u An• Plmlathemiy" 1933, S• 233•

9fiI~lt~mi~~ Vol. 12n No. 5 (1938)

144 T. ]{ITAGA~VA Vnl: XII

and; therfore,

Substituting /-%,,,,~=45,ooocal. in (la), we have Er/'~4,Soocal.-which corresponds

to the energy of the vibrational state of Br< whose vibrational quantum number

z/' is 5. This shows that the Br,_ molecule which has been in the vibrational states

higher than .>"=5 may be excited to become Br,•. And in fact the presence et the molecules in such vibrational states higher than J'=5 is evident from the

appearance of the absorption bands belonging to these states in the absorption

spectrum of bromine vapour (Table IV). With rising temperature the molecules

can advance to higher vibrational states by gaining vibrational energies, and so it is apparent that in the case of higher temperatures such as the flame a number

of molecules in the states higher than J'=5 exist.

b) The excitation energy required to excite the molecule to the vibrational

state v'=15 in the excited state. is

where Ev'=6,ooocal. for v'= I5 from calculation. When we assume that the mean collision energy (E,u„=5,ooocal.) possessed at the temperature of the flame par-

ticipates in the excitation, we have

Q+E„„>E,,,,+L•v'-Ev", :. Ev"Z5,5oocal.

The vibrational state of Ev"=5,5oocal. corresponds to the vibrational quan-tum number .i'=6. Therefore, the Br._ molecules in the vibrational states such

as v"Z6 will be excited to the vibrational state such as T/=t5 in the Ou* state.

The excitation of the bromine molecule in the Rame can 6e explained from the

reaction energy Q=4o,5oxal. generated in reaction (q)"'.

(3) As at the temperature of the flame (about Boo°K) the mean kinetic energy of molecules is only 2~-3 kcal., it can not be considered that the whole of the reaction heat qo.5 kcal. turns into the translational- energy of the molecules

generated, i.c. HBr and Br, but that almost the whole becomes the inner energy of those molecules. I~Br or Br, however, has none of the electronic excited state in the neighbourhood of the energy level. as qo.5 kcal Accordingly, if the rota-

tional energy of HBr is not taken into cosideration, Q is to be accumulated as

rq) in the chemical reaction system of the flame H- and Br-atoms hove of course been generated, lwl in the speclmm of the flame neither line spectm due In these nlonss nor many lined epectmm belonging In the lIS nu9ende could be found. This may be ascribed k, the fact llmt the

excitation energy l/~of Ihese atoms or,molecules.is far higher Ihan the reaction heal Q of reaction (9} 1•'or the appearance of the envision spectra n( FI_; H and Br in We nlsservaLle range they

require at least the following energies: Iiy ;ISknl.: li (Balmer ¢), s76 kcal.: Br, rsoknl.

9fi1~1t~mi~1 Vol. 12n No. 5

No. u. ~ E\11SS10N SPECTKU3i OF '1'[[E tLAME OP BROJiINF, 145

the vibrational energy of HBr immediately after reaction (q). Let HBr* repre-

sent the hydrogen bromide molecule possessing so much vibrational energy. The

calculation made"'shows [flat it is possible- for the normal hydrogen bromide mole-

cule to have- the energy 40.5 kcal. if it is in the vibrational state whose vibrational

quantum number J' is 6. In the studies of the infra-red absorption spectra of hydrogen bromide the vibrational states such as rr'=a, [, o in the '_`' state have been observed and in the near future the presence of such a vibrational state as

;/'=6 will be ascertained.

(.}) Let tts assume that the mean life of HBr° is considerably long and its vibrational energy is scarcely lost before collision. Then we can imagine that in

the collision of HBr" with Br. the: excess energy of HBr' may be transferred to

the collietittg Bry molecule, and the latter may be excited, t}tus:

Br,('£a)+HBr*=Brz*(Ou*)+IIBr. (13)

13r_* fornYed in ([3) will again return to its normal state emitting the band spectrum in the flame, thus:

Brs*(Ou ̀)= I3r_(r S~a) d' hv, (I ¢)

where v is the frequency of the emission band.

In short, the mechanism of the emission of the band spectrum is explained as

follows: the reaction energy D produced in reaction (9) is accumulated in

HBr as the vibrational energy immediately after the reaction, and when it collides with Bre, this molecule may be excited. The excited Br, molecule presents the

band emission while it returns to its normal state. As the energy o(the spectrum

in this case originates from the chemical reaction between bromine and hydrogen,

the emission of the band spectrum of the flame can be regarded as chemilumines-

cence.

(ii) Branching mechanism of the chain reaction

It has been admitted that a combustion or explosion reaction has a mechanism of a branching .chain reaction. Let us consider the branching mechanism of the chain in the combustion of bromine and hydrogen.

In Table V the dissociation energy of Brs is ¢5.? kcal, which is approxi-

mately equal to its excitation energy. Therefore, when I-IBrY* collides with BrY, there occurs not only the excitation (I3) but also the dctontposition of Br,:

Br,("'R)+HBr*=2BrtIIBr. (t5)

i

rg) The talcalation was done according to formula (q) bq applying the molecular constants are =xGg7em-~ and sweeg4em-t of IIItt{~y'} ,

(1938)

9fiI~lt~mi~1 Vol. 12n No.

146, T. XITAGAWA Val. XI!

The Br atom thus generated becomes the origin of a new chain reaction, and thus the chain branches.

.Other ways of branching of the chain reaction may also be considered. For

example, (A) By* has the excess energy just sufficient for the dissociation energy

of the normal Brz molecule as its excitation energy. Therefore, if Rrp* collides with Br_, according to the resonance of energies the latter decomposes thus:

Br_('rR)}Bra*(Ou`)=2Br+By('s`). (16)

(B) The dissociation heat of Bro* being 1o.7kcal. as seen in Table V, it is far more unstahle than the normal molecule, and apt to decompose even by simple

collision, which has been confirmed by W. Jost16/ in his study of the photochemical

reaction between bromine and Irydrogen.

In these nays, the chain reaction branches so frequently that the reaction velocity increases very rapidly and either explosion or com-

Fig: 3. bustion occurs. As the concentration of Brr` is extremely Schemntic F.zpressinn

low compared with that of Br.,, it seems more probable that of the Branching of the

the chain reaction branches according to the scheme as (3)- Br.•IL-Chain Renctinn.

(9)-(t5), which is shown in Fi BC g 3 Which of the two reactions, (!g) or(t5), has lamer proba-

bility to take place can not be easily discussed here. If H

we assume, however, drat both reactions, (13) and (15), gB~\Br proceed at the same time, it is considered [hat the former ~+BYs explains the mechanism of the band emission observed in the \ flame and the latter that of the branching of the chain Br Br

reaction.

Summary.

t) Tn the study of the emission spectrum of the flame of bromine burning in

hydrogen, about 43 emission bands have been found in the range between a.?6575

and 56oot~, and the nave length of the band heads- measured.

2) In comparison of the emission band of the flame with the absorption band

of bromine vapour, both nave lengths have been quite in good agreement,

and so it has been clarified that the emission band in the flame belongs to the

bromine molecule and that itr electron transition is thus: Ou• -~ 'CR.

3) Prom the results obtained by band-head analysis the vibrational quantum

[6) \\'. Jnst, 7,. phrrik. CA<m., 134, gz ([9z8); [B), 3, 95 t[9z9~

5 (1938)

I'AI~1t~~i~1 Vol. 12n No.

No. 5 LM155ION 5FECIRUM OF 'T[ffi FLAME OP III20MINE 147

numbers belonging to the emission bands of bromine have been determined as follows

It has been assured that the excited bromine molecules in the Ou• state are gener-

ated in the chemical reaction system of the flame.

4) The process of the band emission has been explained by the U(r)-Hove of Br_ drawn by Morse's function.

5) The mechanism of the generation of the excited bromine molecules during the combustion reaction has been discussed from the standpoint of the mechanism of chemical reaction and it has been concluded that the band emission of the flame

is a sort of chemiluminescence.

6) The mechanism of the branching of the chain reaction in the case of the

combustion reaction has been proposed thus

(8) Br+H~=HBr+H, ([q) 13r_*-Br_+hv,

(9') H+Bre=HBr'+Br, ([5) IIBr;+Br.,=HBr+2Br,

(t3) H[3r*+Br.,=HRr+Br.*, (t6) Br,*+Br_=Br.+zBr. of these, (8)-(9')-([3)-([¢) represents the mechanism of the band emission and

(8)-(9')-([5)- or (8)-(9'~-{t3)-([6)- the mechanism of the branching of the chain reaction.

In conclusion, the author has great pleasure in expressing his sincere thanks

to Professor S. Horiba and Professor M. Kinmra for their valuable guidance during the course of this work. He also acknowledges his indebtedness to the Japan

Society for the Promotion of Scientific Research for agrant-in-aid.

'L'he laboratory of Phyaioal Cbe»rlalru, Kydto Imperial Unin^raily.

5 (1938)