- Email: [email protected]
Dissociation energies of alkali metal cyanides
Volume ?3, number 3
DISSOCIATIO3J
Juliet
CHEMICAL PHYSICS LETTERS
ENERCLES
N. MULVIHILL
C&rnistr~
Departmeni,
OF ALKALI
METAL
1.5 June 1975
CYANlDES
and Leon F. PHILLIPS LJniversity of Canrerbun:,
Reccivcd 27 Jsnuxry 1975 Revised manuscript rcceivcd 14 Mxch
Chtistchurch,
New Zeala?ld
1975
Equilibrium constants for the reaction 51 + HCN * iMCN + H. where hl is an alkali metal atom have been determined by rnensuring the I:ffect of added CzN, on the intensity of emission of metallic resonance lines from a fuel-lich, pre-mlsed H&+/O:! flame at 1500 K. The yield of HCN from the cyanogen was measured by withdrawing samples through a quartz micro-probe for mass spcctromctric analysis. The cnthalpy change of the reaction was calculated by the third-law method. \Vc find DOh1-c~ to be 497,474,497,487, and 501 kJ mol-’ for ht = Li, NJ, K, Rb, and Cs, respectively, 311* 22 kJ mol-’ .
1. Introduction
In principle tion
the equilibrium
constant
for the reac-
M+HCN=MCN+H can be deterSned uring the proportion
(I) using flame photometry, by measof the metal M remaining as free
atoms in the presence of known concentrations of HCN and H atoms. Measurements of this type were made some time ago for reactions leading to the formation of metallic halides [I]. In a hydrogen-rich flame it is also necessary, in piinciple, to take account of the equilibrium leading to hydroxide formation: M+H20=MOH+H,
(2)
but th$is not difficult provided the stability of the hydroxide is known. in the relatively cool (1.500 K) flame we have used reaction (2) turns out to be ne&ible. Once the eqtiilibrium constant for reaction (l).has been determined it can be combined Njth the partition functions for rextints and products, derived from
.known
the enthalphy
or estimated
spectroscopic
data,
to give
of reaction at absolute zero, and hence the dissociaiion energy of the metal cyanide. The reason that the method ha& not been used previousIy foi-metai cyanides is that HCN is not eri1 tirely stable in the flime @se!, but is gradually con-
verted to water and cvbon oxides, so that the proportion of added CZN2 which is present as HCN at any point in the flame is not easy to predict. In a current study [2] of the breakdown of cpnnogen in the reaction zone of a fuel-rich H2/N7/02 flame we have measured the yield of HCN as a function of the cyanogen flow; the measured HCN concentrations from that study have now been used in the evaluation of the equilibrium constant KLlr.
2. Experimental
The burner and photometric system were as previously described [3]. H atom concentrations were measured by the Li/LiOH method. Meastirements were made with a flame of unburnt composition H,/NZ/02 = 4.51811, and final temperature approximately 1500 K. Metalatom concentrations (z 10e5 atm) were monitored by both emission and absorption spectroscopy; the results PO be discussed here were obtained by emission spectroscopy. The wavelengths observed were: Ll, 670.8 nm; Na, 589.3 nm; K, 769.9 nm; Rb, 780.0 run; Cs, 852.1 nm. Measurements were made 4 mm above the primary reaction zone, in a region where the emission was mainly thermal rather than chemiluminescent. Cyanogen was added to the extent of up to 1% of the unburnt gases; the flame appeared sufficiently unperturbed for useful
Volume 33,number
CHEMICAL PHYSICS LETTERS
3
lSJune1975
From these slopes values of K! have been calculated, -4th results as shown in table 1.
Values of q, the enthalpy change for reaction (1) at absolute zero, have been calculated using partition functions based on the data in the JANAF tables [4]. The parameters for LiCN, RbCN, and CsCN, given in table 2, were estimated by comparison with those for NaCN and KCN. From the A@ values the M-CN dissociation energies have been calculated using a figure of 418 f 4 kJ mol-l for the enthalpy of formation of CNu at 0 K [5]. This corresponds to a value of 499 5 13 kJ for the H-CN dissociation enerB, when we USEthe JANAF figure for the enthalpy of formation of HCN. The limits of error specified for our va!ues of dissociation energy in table 1 are the sum of the above limits for Di_& aqd the limit specified for A$. The values obtained for the dissociation energies agree with those calculated from the JANAF heats of formation of M, CN and MCN, within the combined limits of experimental error. The values of @ calculated from the JAN_4F data do not agree with our experimental values, the discrepancy for both sodium and potassium being outside the combined limits of experimental error. Our stated limits of error in Ah: correspond to an uncertainty of a factor of two in ICI, which seems generous. Presumably this implies the existence of an error in one or more of the enthalpies of formation which contribute to ALL The trends in RI-CN dissociation energy as M varies resemble the trends observed with the fluorides and chlorides [6], although the ener,? range covered by the values of Dt_CN is smaller.
t + +
D-WI
(ahx
IO’)
Fig_ 3. Graphs of (lo/l - 2) W~SLIS [HCN 670.8 nm alld the sodium D lines.
for the lithium
ever, when 1% or more of cyanogen is added to the flame the reaction zone can be seen to lift away from the burner surface. This has the effect of moving a region of more intense emission into the tield of view of the monochromator, which would account for the observed curvature ofthe gaphs in fig. 3. The slor~s of the linear regions of the curves in fig. 3 are given by slope =KI/([H]
fK2 [H20])
.
(3
Table 1 Dntn forwth3lpies of rsaction (1) and of dissociation atm-‘; tip and Di, kJ mol-*) %I
Slope
K1
to bl and CN .‘t ClK ([Hz01
K2
Li
356 64.1 548 272 1000
3.17 0.455 3.89 1.93 7.09 --
a) -ihiswork;
k,”
b) ref. [4].
..
0.011 -
(3) -
259 24 r9 259 12+-g -3 + 9
x
10m3 atm; units: slope,
D”o(M-CN)
d (a)
Na K Xb CS
= 0.16 atm, [H] = 7.1
66 2 i0 70 * 29 -
(4
OJ)
497 2 22
446 + 13 442F31 -
474 -i 22 497 c 22 487 f 22 50! = 22
CHEMICAL
Volume 33, number 3 Table 2 Parramcters used in partition pm; Y, cm-‘)
Function calculations
(units: r,
PHYSICS LETTERS
AF-AFOSR-71-2134 from the United States Air Force Office of Scientific Research.
References Li
Na K Rb cs
116 116 116 116 116
179 199.2 229.4 248 ,268
410 400 370 340 322
267 239 207 202 198
2200 2176 215S 2125 2100
Acknowledgement This work was supported by the New Zealand Universities Research Committee and by Grant
_.
[ I! E.bl. Bulcticz, L-F. Philfips and T.M. Sugdcn, Trans. &Xiday SOC. 57 (1961) 914. [21 J.N. h!~l~~I and LF. Phillips, 15th International Combustion Symposium (The Combustion Institute, Pitt& bur_ph), to bc published. [3j M.J. McEwan and L.F. Philiips, Cambustion Fiame I1 (19672 63. [4 ] JANAF Thermochemical Tables, 2nd Ed., NSRDS-NBS 37 (N&l. Bur. Std., Washington, 1971). [5 J &f-R. Dunn, C.G. Freeman, MA. M&wan kd L.F. Phillips, J. Phys. Chcm. 75 (1971) 2662. [6] A.G. Gaydon, Dissociation cnergics, 2nd Ed. (Chapman and X&, London, 1968).
.‘ f
DISSOCIATIO3J
Juliet
CHEMICAL PHYSICS LETTERS
ENERCLES
N. MULVIHILL
C&rnistr~
Departmeni,
OF ALKALI
METAL
1.5 June 1975
CYANlDES
and Leon F. PHILLIPS LJniversity of Canrerbun:,
Reccivcd 27 Jsnuxry 1975 Revised manuscript rcceivcd 14 Mxch
Chtistchurch,
New Zeala?ld
1975
Equilibrium constants for the reaction 51 + HCN * iMCN + H. where hl is an alkali metal atom have been determined by rnensuring the I:ffect of added CzN, on the intensity of emission of metallic resonance lines from a fuel-lich, pre-mlsed H&+/O:! flame at 1500 K. The yield of HCN from the cyanogen was measured by withdrawing samples through a quartz micro-probe for mass spcctromctric analysis. The cnthalpy change of the reaction was calculated by the third-law method. \Vc find DOh1-c~ to be 497,474,497,487, and 501 kJ mol-’ for ht = Li, NJ, K, Rb, and Cs, respectively, 311* 22 kJ mol-’ .
1. Introduction
In principle tion
the equilibrium
constant
for the reac-
M+HCN=MCN+H can be deterSned uring the proportion
(I) using flame photometry, by measof the metal M remaining as free
atoms in the presence of known concentrations of HCN and H atoms. Measurements of this type were made some time ago for reactions leading to the formation of metallic halides [I]. In a hydrogen-rich flame it is also necessary, in piinciple, to take account of the equilibrium leading to hydroxide formation: M+H20=MOH+H,
(2)
but th$is not difficult provided the stability of the hydroxide is known. in the relatively cool (1.500 K) flame we have used reaction (2) turns out to be ne&ible. Once the eqtiilibrium constant for reaction (l).has been determined it can be combined Njth the partition functions for rextints and products, derived from
.known
the enthalphy
or estimated
spectroscopic
data,
to give
of reaction at absolute zero, and hence the dissociaiion energy of the metal cyanide. The reason that the method ha& not been used previousIy foi-metai cyanides is that HCN is not eri1 tirely stable in the flime @se!, but is gradually con-
verted to water and cvbon oxides, so that the proportion of added CZN2 which is present as HCN at any point in the flame is not easy to predict. In a current study [2] of the breakdown of cpnnogen in the reaction zone of a fuel-rich H2/N7/02 flame we have measured the yield of HCN as a function of the cyanogen flow; the measured HCN concentrations from that study have now been used in the evaluation of the equilibrium constant KLlr.
2. Experimental
The burner and photometric system were as previously described [3]. H atom concentrations were measured by the Li/LiOH method. Meastirements were made with a flame of unburnt composition H,/NZ/02 = 4.51811, and final temperature approximately 1500 K. Metalatom concentrations (z 10e5 atm) were monitored by both emission and absorption spectroscopy; the results PO be discussed here were obtained by emission spectroscopy. The wavelengths observed were: Ll, 670.8 nm; Na, 589.3 nm; K, 769.9 nm; Rb, 780.0 run; Cs, 852.1 nm. Measurements were made 4 mm above the primary reaction zone, in a region where the emission was mainly thermal rather than chemiluminescent. Cyanogen was added to the extent of up to 1% of the unburnt gases; the flame appeared sufficiently unperturbed for useful
Volume 33,number
CHEMICAL PHYSICS LETTERS
3
lSJune1975
From these slopes values of K! have been calculated, -4th results as shown in table 1.
Values of q, the enthalpy change for reaction (1) at absolute zero, have been calculated using partition functions based on the data in the JANAF tables [4]. The parameters for LiCN, RbCN, and CsCN, given in table 2, were estimated by comparison with those for NaCN and KCN. From the A@ values the M-CN dissociation energies have been calculated using a figure of 418 f 4 kJ mol-l for the enthalpy of formation of CNu at 0 K [5]. This corresponds to a value of 499 5 13 kJ for the H-CN dissociation enerB, when we USEthe JANAF figure for the enthalpy of formation of HCN. The limits of error specified for our va!ues of dissociation energy in table 1 are the sum of the above limits for Di_& aqd the limit specified for A$. The values obtained for the dissociation energies agree with those calculated from the JANAF heats of formation of M, CN and MCN, within the combined limits of experimental error. The values of @ calculated from the JAN_4F data do not agree with our experimental values, the discrepancy for both sodium and potassium being outside the combined limits of experimental error. Our stated limits of error in Ah: correspond to an uncertainty of a factor of two in ICI, which seems generous. Presumably this implies the existence of an error in one or more of the enthalpies of formation which contribute to ALL The trends in RI-CN dissociation energy as M varies resemble the trends observed with the fluorides and chlorides [6], although the ener,? range covered by the values of Dt_CN is smaller.
t + +
D-WI
(ahx
IO’)
Fig_ 3. Graphs of (lo/l - 2) W~SLIS [HCN 670.8 nm alld the sodium D lines.
for the lithium
ever, when 1% or more of cyanogen is added to the flame the reaction zone can be seen to lift away from the burner surface. This has the effect of moving a region of more intense emission into the tield of view of the monochromator, which would account for the observed curvature ofthe gaphs in fig. 3. The slor~s of the linear regions of the curves in fig. 3 are given by slope =KI/([H]
fK2 [H20])
.
(3
Table 1 Dntn forwth3lpies of rsaction (1) and of dissociation atm-‘; tip and Di, kJ mol-*) %I
Slope
K1
to bl and CN .‘t ClK ([Hz01
K2
Li
356 64.1 548 272 1000
3.17 0.455 3.89 1.93 7.09 --
a) -ihiswork;
k,”
b) ref. [4].
..
0.011 -
(3) -
259 24 r9 259 12+-g -3 + 9
x
10m3 atm; units: slope,
D”o(M-CN)
d (a)
Na K Xb CS
= 0.16 atm, [H] = 7.1
66 2 i0 70 * 29 -
(4
OJ)
497 2 22
446 + 13 442F31 -
474 -i 22 497 c 22 487 f 22 50! = 22
CHEMICAL
Volume 33, number 3 Table 2 Parramcters used in partition pm; Y, cm-‘)
Function calculations
(units: r,
PHYSICS LETTERS
AF-AFOSR-71-2134 from the United States Air Force Office of Scientific Research.
References Li
Na K Rb cs
116 116 116 116 116
179 199.2 229.4 248 ,268
410 400 370 340 322
267 239 207 202 198
2200 2176 215S 2125 2100
Acknowledgement This work was supported by the New Zealand Universities Research Committee and by Grant
_.
[ I! E.bl. Bulcticz, L-F. Philfips and T.M. Sugdcn, Trans. &Xiday SOC. 57 (1961) 914. [21 J.N. h!~l~~I and LF. Phillips, 15th International Combustion Symposium (The Combustion Institute, Pitt& bur_ph), to bc published. [3j M.J. McEwan and L.F. Philiips, Cambustion Fiame I1 (19672 63. [4 ] JANAF Thermochemical Tables, 2nd Ed., NSRDS-NBS 37 (N&l. Bur. Std., Washington, 1971). [5 J &f-R. Dunn, C.G. Freeman, MA. M&wan kd L.F. Phillips, J. Phys. Chcm. 75 (1971) 2662. [6] A.G. Gaydon, Dissociation cnergics, 2nd Ed. (Chapman and X&, London, 1968).
.‘ f