• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Introduction
 Flame temperatures
 Atomization efficiency
 Choice of optimum experimental...
 Reference
 Appendix
 Biographical sketch
 Copyright














Title: quantitative investigation of the alkali metals in atomic emission and atomic absorption spectroscopy.
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091578/00001
 Material Information
Title: quantitative investigation of the alkali metals in atomic emission and atomic absorption spectroscopy.
Series Title: quantitative investigation of the alkali metals in atomic emission and atomic absorption spectroscopy.
Physical Description: Book
Creator: Mansfield, Clifton Tyler,
 Record Information
Bibliographic ID: UF00091578
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000423991
oclc - 11062502

Downloads

This item has the following downloads:

Binder1 ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
    List of Tables
        Page iv
        Page v
    List of Figures
        Page vi
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
    Flame temperatures
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
    Atomization efficiency
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
    Choice of optimum experimental conditions
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
    Reference
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
    Appendix
        Page 121
        Page 122
        Page 123
    Biographical sketch
        Page 124
        Page 125
    Copyright
        Copyright
Full Text












A QUANTITATIVE INVESTIGATION OF THE

ALKALI METALS IN ATOMIC EMISSION AND

ATOMIC ABSORPTION SPECTROSCOPY










By
CLIFTON TYLER MANSFIELD


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL 01
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY










UNIVERSITY OF FLORIDA
August, 1963












ACKNOMLEDGEMENTS


The author wishes to express his appreciation to the

chairman of his committee, Dr. J. D. Winefordner, for his

assistance in the preparation of this paper and his daily

encouragement. He would also like to thank the other members

of his committee. The perseverance and companionship of his

wife, Nancy, are recognized and valued.












TABLE OF CONTENTS



ACKNOWLEDGEMENTS......................................

LIST OF TABLES........................................

LIST OF FIGURES.......................................

INTRODUCTION ..........................................

CHAPTER

I. FLAME TEMPERATURES...........................

A. Theory.......... ...... ... ... .... .........

B. Experimental ...... ............... ........

C. Results and Discussion ...................

II. ATOMIZATION EFFICIENCY.........................

A. Theory...................................

B. Experimental.. ..........................

C. Results and Discussion...................

III. CHOICE OF OPTIMUM EXPERIMENTAL CONDITIONS....

A. Theory..................................

B. Experimental ............................

C. Results and Conclusions..................

LIST OF REFERENCES....................................

APPENDIX...............................................

BIOGRAPHICAL SKETCH................................


Page

ii

iv

vii

1



12

,13

15

19

27

28

30

34

48

50

67

70

116

121

124












LIST OF TABLES


Table Page

1. Evaluation of Constants in Empirical Equation
for Determination of Temperatures in B2/02
Flames.......................................... 22

2. Evaluation of Constants in'Empirical Equation
for Determination of Temperatures of C2I2/02
Flames......................................... 23

3. Efficiency of Introduction of Aqueous Solutions
into H2/02 Flames when Using Total Consumption
Atomizer Burners in Flame Photometry............ 43

4. Efficiency of Introduction of Aqueous Solutions
into C2H2/02 Flames when Using Total Consumption
Atomizer Burners in Flama Photometry............ 45

5. Efficiency of Introduction of Mlethanol-Water
Solutions into H2/02 Flames when Using Total
Consumption Atomizer Burners in Flame Photometry 46

6. Efficiency of Introduction of Methanol-Water
Solutions into C2H2/0' Flames when Using Total
Consumption Atomizer Burners in Flame Photometry 47

7. Spectral Data .................................. 55

8. Conversion of Solution Concentration in Moles/L.
to Atoms/Cm. in Flame Gases for Oxygen-Hydrogen
Flames............................................ 65

9. Conversion of Solution Concentration in Moles/L.
to Atoms/Cm.3 in Flame Gases for Oxygen-
Acetylene Flames................................ 66

10.-15. Sodium in the Hydrogen Flame at 5890 A.......... '81-86

16.-21. Sodium in the Acetylene Flame at 5890 A......... 87-92

22.-26. Potassium in the Hydrogen Flame at 7665 A....... 93-97

27.-30. Potassium in the Acetylene Flame at 7665 A...... 98-101

31.-34.. Potassium in the Hydrogen Flame at 4045 A....... 102-105









35.-38. Potassium in the Acetylene Flame at 4045 A...... 106-109

39. Sodium in the Meker Burner Flame at 5890 A...... 110

40. Potassium in the Meker Burner Flame at 7665 A... 110

41. Atomic Absorbance of Na 5890 A. Line in H2/02
Flames.......................................... 111

42. Atomic Absorbance of the Na 5890 A. Line in
C2H2/02 Flames ................................. 112

43. Atomic Absorption of K 7665 A. Line in H2/02
Flames.......................................... 113

44. Atomic Absorption of K 7665 A. Line in C2BH/02
Flames.......................................... 114

45. Atomic Absorption of Na 5890 A. Line in the
Meker Burner Flame.............................. 115

46. Atomic Absorption of K 7665 A. Line in the
Maker Burner Flame.............................. 115













LIST OF FIGURES


Figure Page

1. Emission Analysis Equipment..................... 18

2. Atomic Absorption Apparatus.................... 32

3. Atomic Absorption Apparatus (photograph)........ 33

4. Cross-sectional Distribution in a Flame......... 36

5. Cross-sectional Distribution in a Flame......... 37

6. Example of Curves by Smit and Vendrik........... 49

7. Sequence of Events Occuring on Introduction of
a Sample Solution into a Flame................... 51

8. Calculated N-T Curves for Na 5890 A. Line....... 61

9. Calculated N-T Curves for K 7665 A. Line........ .62

10. Calculated N-T Curves for K 4045 A. Line........ 63

11. Chamber-type Atomizer and Burner................. 69

12. Example of a Working Curve for a C2H2/02 Flame.. 71

13. Example of a Working Curve for a H2g/0 Flame.... 72

14. Experimental Limits for Na 5890 A. Line in H2/02
Flame........................................... 74

15. Experimental Limits for Na 5890 A. Line in
C2H2/02 Flames.................................. 75

16. Variation of Na Cl Dissociation Constant with
Temperature..................................... 122

17. Variation of KC1 Dissociation Constant with
Temperature ............*. ...................... 123















INTRODUCTION


Flame photometry is well established as a spectrochemi-

cal method of analysis. It is the most common method for

the analysis of alkali metals, and it is quite popular for

the analysis of the alkaline earth metals and some of the more

easily excited transition metals.. Approximately forty elements

can be determined quantitatively by flame photometry, although

Gilbert (34) has given spectral data for a total of sixty-

six elements.

Flame photometry had its beginning with the early flame

tests used for the alkali metals and alkaline earth metals.

The discovery of cesium in 1860 and of rubidium in 1861 was

made by Bunsen and Kirchhoff through observation of their

flame spectra (23). The first instrument to use a flame for

quantitative measurements was constructed by Champion, Pellet,

and Grenier (17) in 1873 for the analysis of sodium in plant

tissue. This instrmuent used a visual detection method to

compare the intensity of a flame containing an unknown amount

of sodium to the intensity of a flame containing a known amount

of sodium. Other flame photomaters were later constructed

which used photographic plates to measure light intensity.










Flame photometry was not extensively used by chemists

until Lundegardh, a Swedish agronomist, developed a burner-

atomizer system which simplified sample introduction into

the flame (23). The solution was sprayed from an atomizer

into a chamber under controlled conditions. Then, the mist

vas fed into the base of an air-acetylene burner and into

;he flame. A quartz-prism spectrograph served as the mono-

chromator, and the emitted radiation was recorded on a photo-

graphic plate. After this development by Lundegardh, a num-

ber of European chemists began to utilize flame photometry.

Among those taking part in this development were Jansen, Beyes,

and Richter (42) and Schuhknecht (58). Siemans and Zeiss

marketed commercial instruments based on principles developed

by Schuhknecht. These instruments used a prism monochromator

and a photocell and galvanom=ter for the detection system.

It was not until the Lundegardh method was introduced

in 1939 by Griggs (35) and Ells and Marshall (23) that flame

photometry became a popular method in the United States. Further

work by Barnes ge al. (9) led to the development of the first

commercial model flame photometer marketed in this country,

the Perkin Elmer Nodel 18. This early model was a filter

instrument which has since been replaced by a dual-optical

path prism system. In 1948, Beckman Instruments made available

a flame attachment for their Littrow prism instruments (23).










Latter Models of this attachment combined with the Beckman

DU spectrophotometer are the most widely used flame photometers

at present.

The development of the multiplier phototube and the re-

duction in cost of fine gratings have given flame photometry

the sensitivity and selectivity needed. Although sodium and

potassium have been the primary elements determined by this

method, increased speed and accuracy of measurements are ex-

tending its use to other elements where more troublesome methods

have been used. A number of books cover the general uses of

flame photometry (16, 23, 36, 48, 53, 59).

A closely related method that has recently extended the

range of elements for which flame methods are applicable is

atomic absorption spectroscopy. In atomic absorption spec-

troscopy the flame serves primarily to vaporize the solution

and dissociate the compounds. A beam of radiation which is

characteristic of the element to be analyzed (usually one of

the resonance lines) is passed through the flame and into a

monochromator. The intensity of the light is measured with

and without solution being atomized into the flame, and the

amount of light absorbed with solution in the flame is a func-

tion of the concentration of the element in the solution.

If the line from the light source is narrower than the line

from the flame, the monochromator only serves to separate the

line being measured from the other lines from the light source

or flame.









Although the principle of atomic absorption has long

been known and applied by astrophysicists to the analysis

of stellar atmospheres (27. 57), the method was not widely

used by chemists until articles suggesting its application

were published in 1955 by Walsh (65) and Alkemade and Milatz

(2). The analysis of mercury vapor in the atmosphere was

the only analytical application of atomic absorption spectros-

copy. Since Walsh's initial article, over 40 articles on

applications of atomic absorption have been published. Among

these are several review articles (7, 22, 29, 46, 49, 56, 66,
67) and one book (25). Atomic absorption spectroscopy supple-.

ments flame photometry in that elements which are difficult

or impossible to excite to emission in the flame can be de-

termined. These include Au, Pt, Pd, Rh (47), Ni, Co (5),

Ag (54), Pb (26, 55), Zn (6, 21, 32), and Cd (29).

One is justified in discussing atomic absorption and

atomic emission spectroscopy together in that both methods

make use of an atomizer-burner to vaporize the solution, evap-

orate the solvent from the compounds, and dissociate the com-

pounds into elements. They are equally affected by problems

such as ionization and compound formation of the element to

be analyzed, as will be discussed later. In flame photometry,

an additional step is required in that the elements to be

analyzed must be excited prior to the emission of character-

istic radiation. In both procedures, similar equipment is










used. Only a light source which emits radiation character-

istic of the element to be analyzed is needed to convert a

flame photometer to an atomic absorption spectrophotometer.

At this point the differences in basic theory should be

emphasized. At low concentrations, if self absorption and

chemiluminescence are neglected, the integrated intensity

of an emission line is given by (1, 37, 48, 67)



I)d) = CAN.hi (1)


where A is the Einstein transition probability for spontaneous

emission of the line, Nj is the number of atoms in the excited

state involved in the transition responsible for the line,

h is the Planck constant, J is the line frequency, and C

is a proportionality constant which depends on the instrument

used and on the half-intensity width of the spectral line.

The number of atoms in the excited state is given by the

Boltzman equation,


N N e-'EJ/kT (2)
g

assuming that the part of the flame under observation is in

thermal equilibrium. Here, N represents the number of atoms

in the ground state per cc. of flame gas, g, and g are the

statistical weights for the excited and ground states respec-

tively, E. is the energy involved in transition from the ground










state to the upper state j, k is the Boltzman constant, and

T is the equilibrium temperature. Combining Equations 1 and

2, the equation for the integrated intensity (1, 48, 67) of

a resonance line is


f l d) =Ch)AN g e"E kT (3)
g

From this equation, it can be seen that the emitted intensity

depends on T and E.. The energy, Ej, is inversely proportional

to the wavelength of the line, as the above equation is con-

cerned with resonance lines.

In atomic absorption, a parallel beam of radiation of

intensity Io is passed through the flame, and an atomic ab-

sorption coefficient K at frequency 0 can be defined (67)

by

-K)L
I = oe (4)

where I is the intensity of the transmitted beam of radiation

and L is the path length through the flame in cm. The rela-

tionship between the absorption coefficient and concentration

(67) is given by
2
Kd) = Ir N)f (5)
mc

where e is the electronic charge, m is the electronic mass,

c is the velocity of light, No is the number of atoms per

cc. which are capable of absorbing radiation in the range










3 tobd), and f is the oscillator strength. Because the num-

ber of atoms in the excited states is negligible when com-

pared to the number of atoms in the ground state, No may be

replaced by N, the total number of atoms per cc. of flame

gas.

It can be seen, then, that the temperature dependance

of atomic absorption is very slight when compared with atomic

emission. Atomic absorption only depends on the excitation

potential in the selection of a line for absorption. It is

independent of the excitation potential as far as excitation

in the flame is concerned.

In atomic emission or atomic absorption flame photometry,

it is virtually impossible to try all combinations of instru-

mental and chemical variables. Usually the conditions for

analysis are chosen at random, so it is not surprising that

in many instances the experimental conditions deviate greatly

from the optimum ones. In order to simplify the selection

of optimum or near optimum experimental conditions, it is

necessary that a theoretical and an experimental study be per-

formed in which all significant factors and their interrela-

tionships are considered. In this dissertation an optimum

range of experimental conditions is predicted for the analy-

sis of sodium and potassium by considering instrumental fac-

tors, flame conditions, and flame equilibria.










When working at high flame temperatures in atomic emission

or atomic absorption flame photoe.try, ionization will be

significant (e.g., the alkali metals), and deviation from

working curve linearity will be pronounced at low sample con-

centrations. However, if the flee temperature is too low,

compound dissociation (e.g., the oxides of the alkaline earth

metals) will be incomplete, and deviation from linearity of

the working curve v;ill result at high concentrations. In

addition, in atomic emission and atomic absorption flaa- pho-

tc=atry there is a concentration of sample belt= which deteo-

tion is impossible. As stated previously, Equation 3 for

emission holds only for lc concentrations. At high concen-

trations the integrated intensity varies with the trell-kncwn

square root of concentration law (1). This is due to self

absorption of the emitted radiation. In atomic absorption,

the absorption line profile is unicportcat as long as the

absorption line width is greater than the source line width.

This is generally true when using a hollow cathode discharge

tube as a line source. If an optim. range of experimental

conditions is to be chosen, all of the above factors oust

be considered.

Each of the above factors has been previously studied

by numerous investigators. The books by Dean (23), Herrmann

and AlkcMadc (36), and Mavrodineanu (48) give discussions

of topics such as metal vapor ionization, compound formation










and dissociation in the flame, self absorption of radiation,

sample introduction, and flame characteristics. In the disser-

tation by Alkemade (1) each of the following was performed:

a comprehensive study of instrumental factors, flame compo-

sition and temperature, ionization, compound dissociation,

and self absorption of radiation. Experimental conditions

were chosen in order that each of the mentioned factors could

be extensively studied. A comprehensive paper was given by

Alkemade (3) at the 1956 Spectroscopy Colloquium in which

the factors affecting the emission characteristics of various

metals in a number of flames were qualitatively discussed.

Sobolev (63) has considered the broadening of spectral lines

in flames and arcs as a result of the Doppler and Lorentz

effects, and Kolb and Streed (44) have studied the intensity

of spectral lines originating in flames as a function of self

absorption. At the 1962 Spectroscopy Conference, Alkemade

(4) gave a complete paper concerning the excitation of metal

atoms in flames. Alkemade considered several non-thermal

mechanisms in addition to thermal excitation. More recently

Gibson, Grossman, and Cooke (31) have considered the excitc-

tion processes in flames. They were primarily concerned with

how temperature differences influenced the emission enhance-

ment of organic solvents.

Detailed studies in atomic absorption spectroscopy have










been much less abundant. The factors of importance have been

listed by Walsh (65) and Eltell and Gidley (25). The effect

of line profile on atomis absorption analysis has been studied

by Shimazu and Hashimoto (60). Studies on factors which affect

the concentration in the ground state in flame emission pho-

tometry would apply equally well to atomic absorption spec-

troscopy.

Because most studies in atomic absorption and atomic

emission spectroscopy have covered only one process per study

and have not allowed an expericmnter to choose the over-all

conditions for analysis, the purpose of the present study is

to use the results of previous studies, to correlate them,

and to derive curves, called N-T curves (plotting number of

atoms per cc. of flamc gas versus absolute temperature of the

fla=e). These curves will be used to predict an optimum range

of experimental conditions for analysis by atomic emission

and atomic absorption flame photometry. Use of the derived

curves should require no special skill and should allow the

average laboratory technician to choose optimum conditions

for analysis. Sodium and potassium were chosen for this study

because they are not susceptible to side reactions in the

flame, and sufficient spectral data are available.

In order to derive the N-T curves for sodium and potassium,

it was necessary to maie two initial experimental studies.

These studies involved the variation in flame temperature










with flame composition, solution flow rate into the flame, and

height of the region studied above the inner cone of the flame

and a study of efficiency of sample introduction into the

flame. Because both of these studies are of interest in them-

selves, they are discussed as separate parts of this disser-

tation. The final part of the dissertation is concerned with

the selection of optirm experimental conditions for the quan-

titative analysis of sodium and potassium by atomic emission

and atomic absorption flame photometry.














I. FLAE TEMPERATURES


It is generally agreed that the outer cone of a flame

is in thermal equilibrium (43, 48) and that the most impor-

tant parameter describing this outer cone as a thermal source

is its temperature. As shown by Equation 3, the integrated

intensity of a spectral line varies exponentially with tem-

perature. The absorption of radiation is nearly independent

of temperature, as can be seen from Equations 4 and 5. How-

ever, in both atomic absorption and atomic emission the tem-

perature influences the evaporation of the solvent from the

atomized droplets to produce a salt mist, the dissociation

of the salt into atoms, the ionization of the atoms, and the

association of the atoms with other species in the flame.

The flames usually used in atomic absorption and atomic

emission flame photometry are the H2/02 and C2H2/02 flames.

The most commonly used flames of these gases have been selec-

ted for study. The temperatures of these flames were measured

as functions of flow rates of fuel and oxygen into the flame,

flow rate of solution into the flame, and height above the

inner cone of the flame.

The method used was based on the Ornstein two line method

similar to that described by Broida and Shuler (15). The

temperature was found by a simple calculation after measure-








meat of the relative intensities of two iron lines, Fe 3737.13

A. and Fe 3734.87 A. The temperature measured by this method

is an electronic temperature. This is important, as flame

photometry involves a study of electronic deactivation of

excited atoms or electronic activation of atoms from lower

to higher electronic states. Therefore, an electronic tem-

perature is desired. Gaydon and Wolfhard (30), Broida (14),

and others (11) have described other methods that can be applied.

However, they are either complex or demand special equipment

not normally available in the analytical laboratory.



A. Theory

The expression for the integrated intensity of a spectral
line was given by Equation 3. The temperature of the flame

could be calculated from this equation, if the absolute in-

tensity of a given spectral line were measured, and if the

spectral constants for the line were known. However, the

measurement of absolute intensity of a spectral line is not

an easy process, as it involves calibration of the instrument

used with a standard source. It is much simpler to measure

the relative integrated intensities of two spectral lines;

then, by taking their ratio,

JlI dO Ch) AN(gj/g)eEk (6)

I d) Ch)' N(g'/g)e I









one no longer has to worry about the constant C, as the equa-

tion (15) for the ratio reduces to


fI) d) i) Agje EJ/kT
I, d g ge ,

assuming that both transitions have the same ground state.

Therefore, if the spectral constants are known, the electronic

temperature can be calculated from Equation 7 by measuring

the relative intensities of two spectral lines.

Iron was the element chosen for use with this method

because it emits a number of spectral lines in the flame for

which spectral data are available (20). The choice of a pair

of lines for this method is governed by a number of criteria.

The two lines must be negligibly self absorbed or the self

absorption must be approximately the same for both lines.

Crosswhite (20) has indicated that the two lines chosen, Fe

3737.12 A. and Fe 3734.84 A., show negligible self absorption.

Broida and Lalos (13) have discussed the selection of spec-

tral lines with respect to the values of the transition proba-

bilities and statistical weights. A number of instrumental

factors limit the choice of a line pair. The lines must be

close enough together that the two lines can be scanned in a

short time and at such a speed that the recorder responds

accurately to both lines. The two lines must be close enough

together that the photocathode of the multiplier phototube

has essentially the same response to both lines, or else the

spectral response must be corrected. Finally, the two lines










must be of sufficient intensity as to be accurately measurable

with the equipment used. The two lines selected meet the

above criteria.




B. Experimental


A Jarrel-Ash 13odel 8200 scanning spectrometer (Jarrell-

Ash Co., Newtonville 60, Mass.) which uses an Ebert grating

mounting of 0.5 mater focal length and a grating of 12000

groves per cm., blazed for 5000 A., was used for all measure-

meats. The spectrom3ter was equipped with the manufacturer's

support base, optical bench, and vertical detector housing.

The detector housing was rewired to adapt to the Uodel ph-200

Eldorado universal photomultiplier photometer circuit (Eldorado

Electronics Co., Berkley, Calif.). A 12.5 myv. strip chart

recorder was connected to the output of the photometer cir-

cuit. A variety of Beckman total consumption atomizer burners

(23) (No. 4050 small bore capillary, No. 4020 medium bore

capillary, and No. 4060 large bore capillary with either the

E2 or C2H2 burner jackets) (Beckman Instru=ent Co., Fullerton,

Calif.) could be positioned in front of the entrance slit of

the monochromator. The burners were -cunted on a rod held in

a rider base on the optical bench. The rod was marked in

increments of 0.5 cm. so that the burner could be accurately










raised or lowered in such a manner as to study various regions

of the flace. The flame was placed 2 cm. from the spectro-

meter entrance slit. A flat black baffle with an opening

0.5 cm. by 0.5 cm. was centered and permanently positioned

between the entrance slit and the flame. It was attached

to the sa e rider base as the burner.

The oxygen flow rate (( 02) and the hydrogen or acety-

lene flow rates (0 2 or 0 C2E2) were controlled by single

stage regulators on the tanks and by the Beclban pressure

regulator for the Bec"uan DD flara photo=mter (this consists

of two single stage low pressure regulators mounted in a

cast base). Rotam ters (4. rota=ater tubes, Ace Glass, Inc.,

Vineland, N. J.) were placed in the gas lines between the

Beckman pressure regulator unit and the atomisor burner in

order to accurately read the flc. rates of oxygen and fuel

in cc. per minute. These rotamaters were calibrated for the

range of flcw rates used for each gas by means of water dis-

placement.

The flow rate of solution was controlled by a method

similar to that described by Foster and EuHe (28). The flow

rate of solution (Osoln) into the flame was controlled by

applying pressure to a 50 cc. side-arm flask containing the

solution. A tank of nitrogen was used as a source of pressure.

The pressure was controlled by a single stage regulator on










the tank and another single stage regulator in the line be-

fore the flask. Fine control of solution flc rate was

achieved by inclusion of a needle valve in the line prior to

the solution flask and by a pinch clamp on the line between

the atomizer burner and the flask. The flow rate of solution

was measured by an Ace 1A rotaz-ter tube which was placed

between the atomicer capillary and the flask. Figure 1 shows

a block diagram of the flcw rate control system.

A stock solution of ferric chloride containing 500 p.p.ma.

of iron was prepared for use in all temperature measurements.

This concentration of iron was found to give sufficient inten-

sity for the two lines measured under all conditions selected.

No measurable self-absorption was found for concentrations of

iron of 2000 p.p.m. or below.

To make a temperature ceasure=mnt, the flcw rates of

oxygen and of hydrogen or acetylene and the flow rate of solu-

tion into the fla.m were adjusted to the desired values, and

the values were recorded. The two iron lines were scanned

manually, and the sensitivity of the detector circuit was

adjusted so that the Fe 3737.13 A. line (the more intense of

the pair) was about 80 to 90% of full scale on the recorder.

Then, the two iron lines were scanned automatically at a rate

of 2 A. per minute, and the output signal was recorded. The

scan was repeated at least twice for each temperature measure-

ment. The average ratio of the heights of the two peaks was




















H -





U-I-
Fm


SB


I oxfuel

- ~---- oxygen


I
1<-- nitrogen


Figure 1.
Emission Analysis Equipment.


spectrometer entrance slit
baffle slit
typical flame
Beckman burner
Brooks 1A-15-1 rotameter
Brooks 4-15-2 rotameter
needle valve
50 cc. side-arm flask
single stage regulator
Beckman regulator
=TT= =I---


o0 0


&&










determined, and the flame temperature, T, was calculated using

Equation 7.

An error of one recorder scale division in measuring

the height of either peak (200 divisions represents full

scale) resulted in a 5 error in temperature. Wlith the ex-

perimental setup used, it was found that for a given set of

conditions the relative standard deviation in temperature

measurements was about 5%. Minor fluctuations in solution

flow rate may cause appreciable changes in intensity but only

small changes in the ratio of intensities and T.




C. Results and Discussion


Using the two line method, flame temperatures were mea-

sured as a function of solution flow rate for a large number

of H2/02 and C2E2/02 flames at various ratios of oxygen flow

rate to fuel flcw rate, at various absolute values of Q02,

and at various heights, h, in cm. above the inner cone of the

flame. The average results of flame temperature were plotted

versus solution flow rate for given oxygen to fuel ratios,

at certain values of 02, and at particular heights, h.

It is unnecessary to give all of the measured data of

flame temperatures in a number of graphs or tables, as the

following empirical equation was found to give B2/02 flame










temperatures with -ean errors from the average measuredd values

of less than 50 K.


~T Ttip ch2
2/02 1 + a(x) + b(X ) K. (8)


tip is the value of the temperature just above the tip of

the inner cone of the flaze. The symbols a, b, and c are

constants vrhich have beon empirically evaluated from the mea-

sured temperature versus solution flow rate curves for a vari-

ety of flame conditions. The values of a, b, c, and Ttip

for several fuel rich, for a stoichiometric, and for one oxy-

gen rich fi L area given in Table 1. Tho term x is the solu-

tion flow rate expressed as number of coles of solution per

mole of oxygen gas introduced into the flame and can be cal-

culated from the following equation



x = 1.25 x 103 (soln (9)
2
The term 9 represents the efficiency of sample introduction

into the flama. This term is further defined and data for

various flames are given in Section II of this dissertation.

The empirical equation listed below was found to give

C2H2/02 flame temperatures with mean errors from the measured









values of less than 50 K.


TC242/02 = -r (10)
1 + floge[1 + a(x) + b(x)J"


The symbols a, b, c, and f are co.. t-ts tiL I :L been em-

pirically determined and are listed in Tablc 2, Qai-a with

Ttip data.

The flamrs listed in Tables 1 and 2 c;oro considered to
be the most important flames when employing total consumption

atomizer burners in analytical emission or absorption flarm

photometry. In the event an analyst wishes to know the tempera-

ture of a flame hose composition is not listed in Tables 1

and 2, an approximate flame temperature can still be found

as long as the ratio of oxygen to fuel is in the ranges listed

in Table 1 (0.2 to 0.6) or Table 2 (1.0 to 3.0). In this case,

the value of ( is estimated from data given in Section II,

and Ttip, a, b, c, and f are estimated from data in Tables

1 and 2. Because the values of X a, b, c, f, and Ttip

do not differ greatly, even when the ratio of oxygen to fuel

is varied, estimates of flame temperatures with mean errors

less than 100 K. should be possible.

The denominator of the empirical equation for a H2/02

flame is quite similar to the denominator of the theoretically

derived equation given by Balker and Vallee (8) for a stoichio-

metric H2/02 flame. The denominator of the empirical equation

for a C2B2/02 flame is similar to the denominator of the theo-

retically derived equation by Baker and Vallee (8) for a stoi-





















TABLE 1

EVALUATION OF CONSTANTS IN EMPIRICAL
EQUATION FOR DETERMINATION
OF TEMPERATURES OF H2/02 FLAMES


Flame Conditions
( 02 O 02, co.
7-2 per min.


2500

2500


2000-3500


Ttip, K.

2700

2750


2750


Constants


0.05

0.07

0.07'


0.6 2500


2630 0.09


0.05 40


0.2

0.4

0.5


0.00

0.025

0.025




















TABLE 2

EVALUATION OF CONSTANTS IN EMPIRICAL
EQUATION FOR DETERMINATION
OF TEMPERATURES OF C2H2/02 FLAMES


Flame Conditions
$ 02 @ 02, cc.
SU2n2 per min.


Constants


a D


2500


2000-3000

4000

3000-4500


3.0 4000


3050

3050

3050

3050

3050


5 10 70


5 10


0.04

0.05


5 10 70 0.08

5 10 70 0.08

7 20 70 0.08


1.0

1.5
2.0

2.5


,tip, &-










chiometric (CN)2/02 flame. The numerators of the empirically

derived expressions differ from those predicted by Broida and

Lalos (13). They found that the flame temperature of a stoi-

chiometric H2/02 flams and of a stoichiometric C2H2/02 flame

decreased linearly with height above the inner cone. They

used a 1 umm. high by 0.01 =mm. wide slit and essentially mea-

sured an adiabatic temperature, i.e., the flame temperature

over the height and width of the slit was approximately con-

stant. In this study a baffle 5 cm. high by 5 =m. wide was

used in order to approximate actual flame photometric condi-

tions. When using such a baffle, the flacm temperature with-

in the baffle mill vary both vertically and laterally as pointed

out by Broida and Lalos (13), and an average flame tempera-

ture rather than a true adiabatic temperature is measured.

The use of an outer flams sheath would minimize the lateral

variation and, to scme extent, the vertical variation. How-

ever, most total consumption burners are used for analytical

purposes without the use of an outer flame shield, and so

the measurements in this paper were made without the use of

one. It was empirically determined that the average flame

temperature measured by the two line rcthod decreased with

the square of the height above the inner cone. The use of

an outer flame shield and/or the use of a much smaller baffle

would probably result in the linear decrease of temperature

with height above the inner cone. In most H2/02 flames studied










the inner cone tip was about 5 mm. above the burner tip, and

in most CH2/02 flames used, the inner cone tip was about 10

n. above the burner tip.

The values of the tip temperatures, Ttip, were found

by extrapolation to zero height. The values found compare

quite favorably with similar values given by Broida and Lalos

(13).

The empirical equations give accurate values of tempera-

ture when using small, medium, or large bore capillary atomizers.

The differences in flame temperature resulting from the use

of different atomizers can be attributed primarily to the

difference in efficiency of introduction of sample into the

flaze. The use of the d factor seems to account for these

processes.

A number of conclusions result from a study of the m-

pirical equations. In all cases the temperature decreases

with increasing solution flow rate. For most H2/02 flames

the decrease was about 100 0 K. when the solution flow .rate

increased from 0.5 to 3.0 cc. per minute. For C2H2/02 flames

the decrease was about 2000 K. for the same change in solu-

tion flow rate. Both decreases are not nearly so rapid as

predicted by Baker and Vallee (8). This is a result of several

factors. Baker and Vallee assumed 100% efficiency of sample

introduction at all solution flow rates, which is not valid.

Also, the temperatures in this study were average temperatures

and not adiabatic temperatures.










It was found that the temperatures above the inner cones

of fuel rich flames were approximately the same as the tempera-

ture above the inner cone of a stoichiometric flame. This

was observed by Brbida and Lalos (13). This is probably due

to entrainment of air from the atmosphere.

Although direct comparison with temperature data from

other studies is difficult due to lack of information on con-

ditions, the temperatures as determined by the two-line method

in this study seem to agree with temperatures determined by

other methods (e.g., the line-reversal method and rotational

methods) on flames from total consumption atomizer burners (23).

Therefore, the empirical equations should be useful to the

analyst in selecting the desired flame conditions. As was

pointed out previously, knowledge of flame temperature is

especially important when ionization and compound formation

are problems in the flame.














II. ATOMIZATION EFFICIENCY


The overall atomizer efficiency, i.e., the percentage

of the sample in the solution that actually ends up unsolva-

ted in the flame, is an important parameter in flame emission

and absorption spectrometry. When a sample solution is in-

troduced into a flame via a total consa-ption atcmizer burner,

a series of processes occurs involving: mass transfer of

the sample solution from the sample container to the flame

which results in large drops, dispersion of the large drops

into small droplets, evaporation of the solvent from the salt

to produce a mist of salt particles, and dissociation of the

salt into atoms, which depends on the nature of the salt, the

temperature of the flame, and the flame composition. Although

each of these individual processes have been individually

studied (24, 36, 48), no study has been made concerning the

overall atomization efficiency. The complete process for any

given burner and flame composition should be of more impor-

tance to the analyst.

The efficiency of sample introduction, a as used in

this study, will be defined as follows: for any given flame

composition and temperature and for any given region, the

efficiency is the ratio of the concentration of unsolvated

salt experimentally present in the flame, whether in atomic,










molecular, or ionic form, to the concentration of unsolvated

salt which would be present if the processes of dispersion

and solvent evaporation were 100% complete. In this study,

zinc chloride in aqueous and nonaqueous solutions was intro-

duced into a variety of flames, and the absorption of reso-

nance radiation by zinc atoms was used to measure the effic-

iencies of sample introduction as a function of flame temp-

erature, flame composition, region measured, and solvent.

The factors which influence the rate at which a sample

solution flows into a flame when using a total consumption

atomizer burner have been studied by Winefordner and Latz

(68). The process of dispersion of large drops into small

droplets, particularly with regard to the factors controlling

the size of the droplets formed, has been reviewed by Dean

(23) and Herrmann and Alkemade (36). Dean and Carnes (24)

studied the size of droplets formed when using total consump-

tion burners. The process of evaporation of solvent from

the solvated salt particles has been reviewed by Herrmann

and Alkemade (36)



A. Theory

In experimentally determining atomizer efficiency, it

is not practical to measure the concentrations of all forms

of a given salt in a flame. The simplest method is to intro-

duce into a given flame a salt which is essentially completely










dissociated for all flame compositions and temperatures of

interest. In addition, the dissociated metal atoms must not

ionize appreciably at the flame temperature studied, and a

sharp line resonance lamp of the particular element must be

available. A salt which essentially fulfills these criteria

is zinc chloride. Zinc salts, according to Clinton (18), are

completely dissociated even in cool coal gas flames, e.g.,

a Bunsen burner flame. Zinc atoms, even at temperatures up

to 40000 K. and at low concentrations, undergo negligible

ionization because of the high first ionization constant,

i.e., 9.3 ev. (36). Therefore, ZnC12 was used in all effic-

iency measurements in this study.

The method of calculating the efficiency of sample in-

troduction is as follows. A zinc chloride solution is intro-

duced into a flame, in which the composition is known, at a

flow rate of t, in cc. per minute, and the absorbance, A ,

is measured at a selected height above the tip of the inner

cone of the flame. Similarly, the same zinc chloride solu-

tion is introduced into the flame at such a low flow rate,

min, in cc. per minute, that the process of dispersion and

the process of evaporation are approximately complete, and

the absorbance, Amin, is measured. The efficiency of sample

introduction, 5 can be calculated from the following equa-

tion:
S A (11)

Amin( /min)










In all of the measurements perfor ed in this study, Amin was

0.10 cc. per minute, and so Equation 11 can be rewritten as
S" O.. (12)

Amin





B. Experimental


The same basic equipment used in the study described

in Section I was used in this study. Several modifications

were made in the optical system. A zinc hollo. cathode dis-

charge tube (micro-Tek Instruments, Inc., Baton Rouge, La.),

powered by a well regulated DC po.?er supply (0 to 600 volts,

0 to 30 milliamperes range), was counted on a rod, placed

in a rider base on the optical bench, and was optically aligned.

One spherical quartz lens of 100 = focal length was placed

between the zinc tube and the flame, and a second similar

lens was mounted between the flame and the spectrometer en-

trance slit. The first lens and the zinc source were so moun-

ted that a circular image of the cathode 0.5 ca. in diameter

was focused on the center of the flame. The other lens was

adjusted until the zinc radiation from the discharge tube

fully illuminated the spectrometer entrance slit. A flat

black metal plate was placed in front of the entrance slit

as a shutter. The revised optical arrangement is illustrated










in Figure 2, and the equipment is pictured in Figure 3.

A solution containing 50.0 p.p.m. of zinc as zinc chloride

was prepared for use in all absorption measurements.

Approximately 300 volts was required to start the hollcw

cathode discharge tube. After the lamp fired, the voltage

was adjusted until the current remained at 7.0 ma. The current

was readjusted to 7.0 ma. every 10 minutes, and usually three

readjustments ware required for the lamp current to completely

stabilize at this value. The detector end recorder circuits

were turned on and adjusted to the desired sensitivity. The

spectrometer wavelength control was adjusted to the peak wave-

length of the Zn 21S9 A. line and locked into position. The

flame composition was adjusted to the selected value. The

shutter was placed in front of the spectrometer entrance slit,

and the recorder was adjusted to zero. The shutter was opened,

the recorder was adjusted to nearly full-scale, and the re-

corder deflection (1) was measured. Then, the zinc solution

was introduced into the flame at a flow rate of 0.10 cc. per

minute, and the recorder daflection,lmin, was read. Each

set of measurements was done in duplicate. From each set of

values of I1 and Imin and absorbance, ~min, was calculated,

and the average value of Amin determined. The same procedure

was repeated for solution flow rates greater than 0.10 cc.

per minute, and the average absorption, AO, Was calculated

for each solution flow rate. The efficiency Of sample intro-














23 cm. -- 14 cm. 9 cm. 7 cm.







A B B LD

fuel
oxygen --- = C

A hollow cathode discharge tube o

B quartz lens, 10 cm. focal length solution

C Beckman atomizer burner

D spectrometer entrance slit



Figure 2

Atomic Absorption Apparatus


















































Figure 3.

Atomic Absorption Apparatus.












33










duction, S' for various solution flow rates was calculated

by using Equation 12.




C. Results and Discussion


The validity of Equations 11 and 12 depends upon the

width of the zinc resonance line (2139 A.) from the hollow

cathode discharge tube being appreciably narrower than the

zinc absorption line due to the hot atoms in the given flame

(4, 43). The emission line from the zinc source acts as a

slit for the absorption line. If thL emission line is too

wide, then the measured % transmission, i.e., (I/I) x 100%,

will be too large. The effect of slit width on absorption

lines has been discussed by Broderson (12) and inefordner

(69). Because the emission line width of spectral lines from

hollow cathode tubes depends on the tube current, an experi-

mental study was performed in which values of % transmission

were rmasured at various tube currents (4 to 20 ma.) at a

given solution flow rate and at a single fla-- composition

and temperature. For currents above 10 ma., the emission

line width was about the same or greater than the absorption

line width, as the % transmission values increased quite ra-

pidly. For currents below 10 ma. the % transmission approached

a constant value. All currents of 8 ma. and below gave the

same values, and so a current of 7 ma. was used in all studies.










As previously discussed, Clinton (18) noted that zinc

compounds were completely dissociated even in relatively cool

flames. If this is truo, variation of the zinc concentration

should not result in a measurable change in I This was

experimentally verified over the concentration range of 10

to 200 p.p.m. zinc. The 50 p.p.m. zinc solution was chosen

for all experimental efficiency studies because appreciable

absorption occurred at 0.1 cc. par minute,and the absorption

was still measurable at 3 cc. per minute.

The diameter of the beam of radiation from the resonance

source which passed through the flc=a was 0.5 cm., and the

average width of the fleam gases through which the beam passed

was about 0.7 cm. Therefore, in measuring AA and Amin only

a portion of the total atoms introduced into the flame per

unit time were actually in a position to absorb radiation

from the source. As absorbance is a function of concentra-

tion of atoms in the path of the radiation, some idea of the

distribution of atoms in the flame can be obtained by measuring

the absorbance of a narrow beam of radiation passing through

the flame. By a series of measurements of absorbance from

a beam of radiation 0~1 cm. in diameter, the absorbance as

a function of distance, x, from the vertical center of the

flame was determined experimentally for several flame heights,

compositions, and solution flow rates. Examples of the re-

sults may be seen in Figures 4 and 5. The measurements and

the results obtained were quite similar to those obtained by











0.9


0.8-
$ 02 = 2500 cc. per min.

$ 02/$ C2H2 = 2.5
0.7- Ht. in flame = 1.5 cm.




0.6-



0.5-



0.4- .0 s=2.0



0.3 -
$s=0.5


0.2 -

Ss=0.2

0.1-



0.0
0 0.1 0.2 0.3 0.4 0.5

Distance from Center of Flame, cm.

Figure 4.

Cross-sectional Distribution in a Flame.


0.6

















0.7 Ht. in flame = 2.5 cm.



0.6 -



0.5 -



0.4 =1.0 s=2.0



0.3-

s=0.5

0.2



0.1 S=o.2



0.0 I I I I
0.0 0.1 0.2 0.3 0.4 0.5

Distance from Center of Flame, cm.

Figure 5.

Cross-sectional Distribution in:a Flame.










Gibson, Grossman, and Cooke (31). The distribution of absorb-

ance, Ax, as a function of the distance x from the vertical

center of the flame followed approximately a Gaussian distri-

bution.

By determination of a number of Ax vs. s curves for a

number of flame conditions, it was found that the following

empirical equation gave an accurate means of correcting the

measured efficiencies for the change in width of the flame

gases with change in solution flow rate:


{1.06 0.36(x/s)
c 5' l___ -- - min (13)
J"corr e las(1.06 0.36(x/s))} (


where is the efficiency corrected for flame width, 4 ,,

is the efficiency measured according to the method described

in the experimental section, A x is the half maximum width

in cm. of the Ax vs. x distribution, and s is the width of

the beam of resonance radiation in cm. passing through the

flame gases. The upper term in brackets is evaluated for the

minimum solution flow rate, Qmin, and the lower term in brackets

is evaluated for the desired solution flow rate 0. The value

of ., differs significantly from 5.. only for very large

aqueous solution flow rates and for most organic solution

flow rates. If the change in width of the flame gases is

not greater than two-fold when the solution flow rate is in-

creased from 61in to 4, then A x in the above equation can

be replaced with the approximate value W/2, where W is the










width of the flame gases in cm. The relative standard devi-

ation in measuring efficiencies is 8%, and so this approxi-

mation does not introduce significant error.

The validity of Equations 11 and 12 depends on the use

of a minimum flow rate, 1min, which is sufficiently small

that dispersion and evaporation are essentially complete.

A series of efficiency measurements at aqueous solution flow

rates of 0.15 cc. per minute and below gave the same results.

Therefore, all solution flow rates of 0.15 cc. per minute

and lower were assumed to have 100% efficiency of sample in-

troduction. As can be seen from Table 5, organic solution

flow rates significantly greater than 0.15 cc. per minute

resulted in 100% efficiency. However, the minimum solution

flow rate used in all measurements was 0.10 cc. per minute.

It was not necessary to make any correction for reflec-

tion losses of the zinc resonance radiation due to water

droplets in the flame. Reflection losses could cause errors

in the values of 9 because the I readings were made with

no solution flowing into the flame, whereas all other readings

were made with a definite solution flow rate. Erroneously

high g values could result. It was experimentally determined

that the absorbance due to reflection losses was only 0.008

at 0.20 cc. per minute and only 0.02 at 3 cc. per minute.

These values are negligible in comparison with the absorbance

readings due to zinc absorption at the same flow rates.










In Tables 3 and 4, the sample introduction efficiencies,

S are given for various flames as a function of aqueous

solution flow rates. The flame size remained approximately

constant (as determined by a series of photographs for a se-

ries of flow rates) for all solution flow rates listed in

Tables 3 and 4, and so the correction for change in flame

width was unnecessary.

In Table 5, the sample introduction efficiencies, ,

are given as a function of flow rate of several methanol-water

solutions into H2/02 flames. All efficiencies were corrected

for the change in width of flame gases with change in solu-

tion flow rate.

Several interesting conclusions can be drawn from the

results given in Tables 3 and 4. An increase in solution flow

rate always results in a less efficient sample introduction,

and the rate of change of efficiency with flow rate is always

greatest at low flow rates. The efficiency of most 12/02

flames increases with flame height at any given solution flow

rate, whereas the efficiency of the C2H2/02 flames and several

of the H2/02 flames in which the flow rate of oxygen is high,

decreases with flame height at any given solution flow rate.

The increased efficiency of the H2/02 flames with height is

probably due to the additional time allowed for the processes

of dispersion and evaporation. The data in Tables 3 and 4

are insufficient to permit a discussion of the reasons for

the decrease in efficiencies with increasing height for C2H2/02

flames as well as the few H2/02 flames mentioned above.










From the data in Table 5, it is evident that the sample

introduction into H2/02 flames for all methanol-water mixtures

is more efficient than for aqueous solutions under similar

conditions. The decrease in efficiencies with increasing

solution flow rate is much less rapid than with aqueous solu-

tions. The increase in efficiencies with increase in flame

height is also quite pronounced when using methanol-water

mixtures instead of aqueous solutions. Gibson, Grossman, and

Cooke (31) have explained the increase in flame emission in-

tensity for several elements introduced into the flame via

organic solvents as being essentially due to the increased

sample introduction efficiencies and the increase in flame

temperature. The increase in absorption of radiation in atomic

absorption spectroscopy should be due primarily to the increase

in sample introduction efficiency. This would particularly

be the case when using total consumption atomizer burners.

Efficiency data and flame temperature measurements for other

solvent systems may enable the experimenter to calculate the

approximate increase in emission or absorption for a given

element in a given solvent system and flame.

Table 6 gives solution introduction efficiency data for

methanol-water solutions in C2H2/02 flames. These data :are

difficult to explain, as there was no increase in efficiency

noted when using an organic solvent instead of water. The

C2H2/02 flames studied were quite fuel rich to begin with,










so an increase in flame temperature would not be expected

on the addition of an organic solvent. Due to the extreme

turbulence of these flames, it is possible that some of these

measurements were made on the reaction zone of the flame at

higher solution flow rates.

The bore size of the atomizer capillary seems to have

an effect on sample introduction efficiency. For the same

flame conditions and solution flow rate, the medium bore

atomizer gives better efficiency than the small bore atomizer.

As previously noted (683), the geocatry of total consumption

atomizer burners varies from one to another, which could

possibly account for the differences noted.

The precision of the measurements performed in this study

depends on several factors. The relative standard deviation

of setting and determining the oxygen and fuel flow rates

and all solution flow rates greater than 0.5 cc. per minute

was about 5%. Unfortunately, the relative standard deviation

in r-easuring the 0.10 cc. per minute solution flow rate was

about 10%. The relative standard deviation in measuring the

intensities of the zinc resonance radiation was about 1%,

with and without zinc in the flame. By using the 50.0 p.p.m.

zinc solution for all measurements in this study, the rela-

tive standard deviation in log (IOI) did not exceed 10 for

all conditions. Therefore, the standard deviation in all

measurements was about 10%.












Table 3


EFFICIENCY OF INTRODUCTION OF AQUEOUS SOLUTIONS INTO /02
FLAMES WHEN USING TOTAL CONSUr2TION ATOMIZER BURNERS I
FLAME PHOTO'ETRY

(small bore capillary)


0,- 2000 2000 2000 2500 2500 3000 3000 3000
02
0 2 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
2 H2
h, cm. 0.5 1.0 1.5 0.5 1.5 0.5 1.0 1.5



solara /


0.1 1.00 1.00 1.00 1.03 1.00 1.00 1.00 1.00
0.5 0.82 0.81 0.98 0.78 0.74 0.76 0.71 0.73
1.0 0.61 0.77 0.84 0.60 0.56 0.55 0.61 --
2.0 0.43 0.55 0.55 0.48 0.36 0.45 0.57 0.48
3.0 0.39 0.45 0.43 0.43 0.31 0.44 0.54 0.44


0o2 3500 3500 3500 2500 2500 2500

0 o2 / 0.5 0.5 0.5 0.2 0.4 0.6

h, cm. 0.5 1.0 1.5 1.5 1.5 1.5



Osoln


0.1 1.00 1.00 1.00 1.00 1.00 1.00
0.5 0.88 0.84 0.97 0.80 0.93 0.74
1.0 0.74 0.79 0.81 0.60 0.67 0.60
2.0 0.60 0.72 0.65 0.50 0.47 0.45
3.0 0.57 0.58 0.59 0.47 0.45 0.36
J' --= al II i










Table 3 (Cont.)


(medium bore capillary)


602 2500 2500 2500 3000 3000 3000
00 /0 0.5 0.5 0.5 0.5 0.5 0.5
h, cm. 0.5 1.0 1.5 0.5 1.0 1.5



0soln


0.1 1.00 1.00 1.00 1.00 1.00 1.00
0.5 0.94 0.93 0.69 0.46 0.63 0.75
1.0 0.74 0,78 0.62 0.42 0,55 0,61
2.0 0,55 0.54 0.51 0.83 0.41 0.53
3.0 0.47 0.44 0.44 0.31 0.36 0.49

Key to Symbols Used:
02 = flow rate of oxygen in cc. per min.

002 /H2 (or =02 /C2 2) = ratio of oxygen to fuel flow rates

h = height of region measured above the inner cone in cm.

Osola =.solution flow rate in cc. per min.
S= efficiency of sample introduction











Table 4


EFFICIENCY OF INTRODUCTION OF AQUEOUS SOLUTIONS INTO C2H2/O2
FLAMES WHEN USING TOTAL CONSUMPTION ATO:IIZER BURNERS IN
FLAME PHOTOMETRY.*



0 2 2000 2000 2000 2500 2500 2500 3000 3000 3000

40 2' aH 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5
h, cm. 0.5 1.0 1.5 0.5 1.0 1.5 0.5 1.0 1.5


4soln


0.1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.5 0.78 0.68 0.56 0.86 0.62 0.65 0.64 0.57 0.63
1.0 0.61 0.49 0.37 0.63 0.46 0.45 0.47 0.43 0.47
2.0 0.45 0.31 0.21 0.44 0.28 0.26 0.30 0.27 0.28


002 4000 3500 4000 4500 4000

2
%0/0o/ 2.0 2.5 2.5 2.5 3.0
h, cm. 1.0 1.0 1.0 1.0 1.0


soln


0.1 1.00 1.00 1.00 1.00 1.00
0.5 0.87 0.84 0.87 0.87 0.74
1.0 0.75 0.68 0.73 0.77 0.55
2.0 0.60 0.47 0.51 0.53 0.44


*The data in the upper section of this table was obtained
using a different medium bore capillary atomizer burner than
the one used to obtain the data in the lower section.











Table 5


EFFICIENCY OF INTRODUCTION OF LETHANOL-WATER SOLUTIONS INTO
H2/02 FLAMES WHEN USING TOTAL CONSUMPTION ATOMIZER BURNERS
IN FLA=E PHOTOMETRY


(medium bore capillary)


M% ethanol 20 50 80 20 50 80

h, cm. 0.5 0.5. 0.5 1.5 1.5 1.5


*soln


0.1 1.00 1.00 1.00 1.00 1.00 1.00
0.2 0.93 1.03 1.00 0.93 1.00 1.00
0.5 0.87 0.94 1.00 0.81 1.00 1.00
1.0 0.83 0.85 0.86 0.60 0.95 1.00
2.0 0.60 0.64 0.67 0.55 0.84 0.98


02


= 2500 cc. per minute


02 A2


= 0.5











Table 6


EFFICIENCY OF INTRODUCTION OF IET.F NGL-lJATER SOLUTIONS INTO
C2H2/02 FLAMES WHN USING TOTAL CONSUMPTION ATOMIZER BURNERS
IN FLAMP PHOTOi2TRY

(medium bore capillary)

M% ethanol 20 50 80 20 50 80
h, cm. 1.5 1.5 1.5 2.5 2.5 2.5


soln


0.1 1.0 1.0 1.0 1.0 1.0 1.0
0.2 1.0 1.0 0.98 0.85 1.0 0.95
0.5 1.0 1.0 0.95 0.83 1.0 0.89
1.0 1.0 1.0 0.93 0.81 1.0 0.75
2.0 1.0 1.0 0.85 0.51 0.73 0.46


02


= 2500 cc. per minute


o2 /C22 = 1.5















III. CHOICE OF OPTI1UM EXPERIMENTAL CONDITIONS


The choice of optimum exparicmntal conditions in flame

emission and absorption spectroscopy requires the considera-

tion of many factors, and if done empirically, would require

much time and effort. Smit and Vendrik (61), while studying

the effect of flame temperature and concentration of sample

on ionization in the flame, were able to graphically illus-

trate the effect of flame temperature and concentration on

ionization, compound dissociation, self absorption, and limit

of detection. A graph similar to the one given by them for

KI is given in Figure 6. Curve a represents the number of

atoms per cc. of flame gas versus temperature for 90% disso-

ciation of the compound, IX, in the flame. Below this curve,

the compound is essentially completely dissociated. Curve

b represents the number of atoms per cc. of flame gas versus

tomperature for 10A ionization of the atoms, M, in the flame.

Above this curve, very little ionization of LI occurs. Curve

c1 represents the limit of detection by flame emission of

the atoms in the flame by the 15 resonance line. Above this

curve, detection is possible. Curve c2 represents 10% self

absorption of the emitted radiation by the atoms in the flame.

Below this curve, self absorption is negligible. The cross-

































w
Co



co
1-4


0










hatched region represents a range of flame conditions which

are optimum as far as all of the curves are concerned. Smit

and Vendrik made no further attempt to apply this method to

other compounds in the flame. This was probably due to a

lack of spectral data.

In this section such N-T curves are calculated for NaCl

and KC1. Although most of the temperatures of flames from

total consumption atomizer burners lie outside the optimum

regions of the N-T curves, these graphs are still useful in

selecting working conditions in flame emission and absorption

spectrometry and explaining deviations of working curves from

linearity.

Experimentally, measurements of a number of line inten-

sities in flame emission spectrometry and a number of absorb-

ancies in flame absorption spectrometry were made for a large

number of flame conditions and solution concentrations for

sodium and potassium. These are used in conjunction with the

theoretical plots in helping to verify working conditions

for sodium and potassium.





A. Theory


Figure 7 gives the sequence of events occurring on intro-

duction of a sample into a flame. The order of events is









I
Transport of Sample into Flame.


MX nS
sample solution
in cuvette


large droplets
of sample


II
Dispersion of Large


Droplets.


0

small drop-
lets of sample
/


+ III
Evaporation of
S IX Solvent and Subli-
Ionization. mation of Salt.
V IV
M*.----Excitation by Thermal--- M(g)<-- Dissociation of Salt.-MX(g) salt mist
/ M e a n s V I I

VI Compound Formation with Flame Gas Produ
onal Deactivation. \ 'lY(g
X MY(g)


cts.


/ VII
Self Absorption of Photon by M.

M*


Excitation by
Radiation from External Source.


Figure 7.
Sequence of Events Oncuring on Introduction of a Sample Solution into a Flame (MX'nS is
a salt of a cation M and an anion X" and having n molecules of S of solvation).


Radiati
1


- --










not meant to imply a given mechanism or is the lack of arrows

in the reverse direction in processes V, VIII, and IX meant

to indicate a lack of equilibrium. The diagram is used to

indicate the most significant processes which must be considered

in the choice of optimum experimental conditions. Processes

V, VI, and VII are of importance only in atomic emission flame

photometry, and process X is of importance only in atomic ab-

sorption flame photometry.

In order to use the Smit-Vendrik method, the flame re-

gion of interest must be in thermal and chemical equilibrium.

It has been shown (43, 48) that the outer cone of analytical

flames under atmospheric pressure is approximately in thermal

equilibrium, i.e., the outer cone of a flame is an approxi-

mate adiabatic system in which all parts are at the same tem-

perature. A flame in thermal equilibrium is described by a

temperature T which determines the number of atoms in any

excited level according to the Boltzmann equation. For an

extensive discussion of thermal equilibrium in flames and the

reasons for deviation from equilibrium see Alkemade's paper

given at the 1962 Spectroscopy Conference (4). It has been

shown that in the outer cone of analytical flames chemical

equilibrium (4, 30, 48) also approximately exists, i.e., at

any temperature, a constant describes a given system. In

addition, it vill be assumed that the ideal gas law is valid

for species in the flames of interest, as has been shown to

a good approximation (48).










The dissociation of compound MX (process VIII) can be

described by the following equilibrium:



MX M + X Kx =, (14)
p
SMX


where KMX is the equilibrium constant of the process at the

temperature of the flame, and the p's are partial pressures

of the components. If no excess of U or X is introduced into

the flame, then p, = p and

P2
K = -- (15)
MX p
*IMX

If P is the total pressure of M in all forms, i.e., P =

pm + p and if oc, is the degree of compound dissociation,

then

2
K = (16)
MX 1 (16)


Since the ideal gas law is valid (P = NkT, where N is the to-

tal number of atoms of M in all forms, k is the Boltzmann

constant, and I is the flame temperature in K.), then N is

given by


(1 (d)
= a4 (KMXlk)









The atom M could also form stable compounds with flame

gas products, i.e., compounds such as MO, MOH, MH, etc. In

this case, P = PM + pMX + pMO + *.' and additional equilibria

must be considered. A significantly more complex expression

would be needed to describe the relationship between N and

T than Equation 17. Fortunately, compound formation of Na

and K with flame gas products can be neglected (40, 41).

Using statistical thermodynamic methods KMX can be cal-

culated as a function of temperature (see Appendix). Mavro-

dineanu (48) gives an equation which can be used to calculate

equilibrium constants for the dissociation of gaseous diatomic

molecules, e.g., the alkali metal halides. Spectral data

necessary for calculation of KMX can be found from a number

of sources (38, 45, 62). Calculations of equilibrium constants

for the dissociation of polyatomic molecules, e.g., MgC12,

MgOH, etc., are considerably more complex. Moelwyn-Hughes (51)

lists a number of general equations for several types of equi-

libria. Unfortunately, the lack of spectral data often pre-

vents the use of such equations. For the alkali metal halides

of Na and K, polyatomic molecules need not be considered,

and so Equation 17 is quite general for considering compound

dissociation of sodium and potassium halides.

Using Equation 17, it is possible to calculate the maxi-

mum value of N for which the degree of dissociation is a given

value, e.g., o(4 = 0.9, if KMX is known at the desired tempera-

ture. KMX can be calculated by using the equation given













Table 7


SPECTRAL DATA


-1
Atom AI, A. gl gu Au 1, sec Eexc' ev.


Na 5890.0 2 4 1.3 x 108 2.10
Na 5895.9 2 2 1.3 x 108 2.10

K 4044.1 2 4 5.7 x 106 3.59
K 4047.2 2 2 5.7 x 10 3.59
K 7664.9 2 4 3.9 x 107 1.61
K 7699.0 2 2 3.9 x 107 1.61


1st Ionization Potential of Na = 5.12 ev.
1st Ionization Potential of K = 4.32 ev.


Energy of Dissociation of NaCi
Energy of Dissociation of KCI


Spectrometer Slit Height
Spectrometer Slit Width
Spectrometer Speed


= 3.59 ev.
= 4.4 ev.


= 0.5 cm.
= 0.0025 ca.
= 0.1










by Havrodineanu (48) and the data listed in Table 7. A plot

of N versus T for c4 = 0.9 for NaCI is given in Figure 8,

and plots of N versus T for cd = 0.9 for KC1 are given in

Figures 9 and 10. Similar curves can be derived for the

other halide salts of sodium and potassium.

The ionization of metal vapor M can be described by the

following equilibrium:


= + e Ki P ep.e" (18)
PM *

In the outer cone of most flames, especially non-organic flames,

the concentration of free electrons is small (30), and so

P+; Pe-, and

2
Pn+
Ki = -- (19)


Compound formation is usually negligible in cases in which

ionization of matal vapor is appreciable, end ionization is

usually negligible when compound formation is appreciable.

Therefore, the total pressure P of metal atoms in all forms

is approximately given by P = pa + p,, when ionization

is significant. If ac represents the degree of ionization

of M atoms, then



K. = ,, (20)
1 1 oL










Assuming that the ideal gas law is valid, N is given by

1 -o
N = (Ki/kL) (21)



Using statistical thermodynamics, Ki can be calculated

as a function of temperature. Uavrodineanu (43) gives an

equation which can be used to evaluate Ki. Spectral data

necessary for calculating Ei can be found in several sources

(45, 62). From Equation 21, it is possible to calculate the

maxinmu value of N for which the degree of ionization is a

given value. A plot of N versus T, for Pk = 0.1, is given

for No i 114m 8 uad for E i Figures 9 and 10.

The area between the compound dissociation and ioniza-

tion curves in Figure 8 corresponds to a region where ioni-

zation is less than (1Q and compound dissociation is greater

than 9%4. Since both atomic emission and atomic absorption

flame photometry depend on the number of atoms in the ground

state per cc. of fl.- gases, this region should be optimum

for spectral analysis. Eot ever, several factors limit the

portion of this region which can be used.

Iu the case of KCI, such an optimum region is nonozis-

tent, as can be seen from Figures 9 and 10.- The curves for

4 = 0.90 and the curves for d; = 0.10 are, for all prac-

tical purposes, identical. This is because the ionization

potential for K (4.32 ev.) and the energy of dissociation










for KC1 (4.4 ev.) are of similar value. Additional curves

for the degree of dissociation of KC1 equal to 0.50 are given

in Figures 9 and 10.

In both atomic emission and atomic absorption flame pho-

tometry, N-T curves can be drc7wn for the spectral line in

concern which correspond to the limit of detection by these

methods of the element in concern. In each figure there is

a limit of detection curve for atomic emission and one for

atomic absorption fl=az photometry. The limit of detection

curves are constructed using equations previously derived

by VWinefordner and Vickhrs (70). Eaah curve is so drc.-n that

i a coaceatratioa of the o lCs invoct-O tOd is behl.c the

limit o! detection for any given temperature, the instrumen-

tal setup being used can not detect the spectral lftie. The

curves are drawn only for !2/02 flames, as the curves for

CgI2/02 flasm s are similar to and are of only slightly loaer

value as those for H2/02 flames. Such curves can be readily

constructed for other instrumental setups and for other spec-

tral lines.

In atomic emission flcme photometry, the self absorption

of emitted radiation must also be considered. The fraction

of radiation self absorbed can be estimated by use of the

growth curves used in astrophysics (52, 64). Growth curves

for the alkali metals have been given by Jazs and Sugden

(40, 41). Their plots consist of a series of curves of










log (AglFn 2'/2nrA) versus log (Nfbvn" 2/ir)a) for various
"a" values. The term "a" called the absorption parameter,

is defined as a =ffin ,. /&a where &)L is the Lorentz

half-width and &),is the Doppler half-width of the spectral

line. The other terms are defined as follows: A is the to-

tal absorption (so-called by Mitchell and Zemansky) as defined

in Equation 22 for the case of high N values where self absorp-

tion is important, N is the ground state concentration, f

is the oscillator strength for the spectral line, and b is

the width of the flame gases being viewed.

2wre2aNfb
Ag Vmc Tn' (22)

In Equation 22 the terms e, c, and m represent the charge of

an electron, the velocity of light, and the mass of an elec-

tron, respectively. For low concentrations where self absorp-

tion is negligible, Ag is given by


Ag (We /mc)nfb (23)

Once the value of "a" is known, the value of N for which

the degree of self absorption is 0.1 can be found as a func-
tion of T in two ways. Hinnov and Kohn (39) give a large

number of "a" values for several spectral lines of a number

of elements. N can be found graphically from the curves of










James and Sugden (40, 41) by finding the value of A at r.hich

self absorption has resulted in a 10% decrease in the value

of A and then, using Equation 22, N can be calculated as

a function of temperature T. The value of N can also be cal-

culated by equating 0.9.A for a dilute gas to A for a =ore

concentrated gas and then


0.94mecaA a A),
N=0.1 = 7 = 1.1 x 102 (24)
e2fb7r fb

The values of m, c, e, and Tr have been substituted into the

latter part of Equation 24. The term A) can be evaluated

as c.-scibod by Iitchell and Zemansky (50), f can be found

in the literature (19), and "a" values can either be taken

from the paper by Hinnov and Kohn (39) or can be calculated

(50). In Figures 8, 9 and 10 curves of N versus T are given

for the case in which the degree of self absorption is 0.1

for the Na 5890 A. line, the K 7665 A. line, and the K 4045

A. line, respectively.

In the above discussion the excitation of sodium and

potcsskiu is assmued to cause a negligible depletion of ground

state atoms. At temperatures below 40000 K. the fraction

of atoms excited can be shc=n (25,65) to be less than 1%

for the Na 5890 A. line and the K 7665 A. line.

In the optimum regions for analysis by atomic emission

filr.e photometry, atomization is at least 90% complete, self

absorption and ionization are less than 10%, and the spectral

liae is detectable by emission. In the optimum r-gicns for











19 1 od = 0.90

18 2 6; = 0.10

3 kA = 0.10

4 limit of detection by atomic absorption spectrometry
16
5 -.limit of detection by atomic emission
15 spectrometry 1


r-t



0 12

011

f10

,o 4


5
o
-7






4

3


1400 1600 1800 2000 2200 2400 2600 2800
T, oK


Figure 8.

Calculated N-T Curves for Na 5890 A. Line.










1 2 dz = 0.10, a = 0.90
3 d".= 0.10
4 limit of detection by
spectrometry
5 limit of detection by
spectrometry


I
1800


-J I I


I I
2000 2200
T, K


atomic absorption

atomic emission


2400


2600


Figure 9.

Calculated N-T Curves for K 7665 A. Line.


19

18--

17-

16 -

15 -

14 -

13 -


4


5


1600
1600


2800


- -n"~~""""~~"-~""""""-?"""-~-~n~~-I~I---









19-
1 o4 = 0.50
2 c1 = 0.10, cj = 0.90
18 3 do= 0.10
4 limit of detection by atomic absorption
17 spectrometry
5 limit of detection by atomic emission
16 spectrometry 1

15 2

14-

13

12 -
II-4

11 4





o
8

0 7
o

2 6

S 5-

4-

3 ,- I I I -
1600 1800 2000 2200 2400 2600 2800

T, K


Figure 10.

Calculated N T Curves for K 4045 A. Line.




64


analysis by atomic absorption spectroscopy, dissociation is

at least 90% complete, ionization is less than 10%, and the

line is detectable by absorption. Therefore, it is expected

that as long as flame conditions and instrumental factors

are such that the concentration of atoms is within the

optimum region for any given flame temperature, then linear

working curves of maximum slope will result.

As it is not practical to work with. N, the number of

atoms per cc. of flame gas, it is desirable to convert N to

molar concentration. Knowing the flame conditions, this is

possible by an equation given by Winefordner and Vickers (70).

3.3 x lOr22QTnTN
C moles/I., (25)
In298 C


where Q is the flow rate of unburned gases in cc./sec. at

room temperature and atmospheric pressure, I is the solution

flow rate into the flame in cc./min., nT is the number of

moles of combustion products at the flame temperature T, n298

is the number of moles of flame gas products which would be

produced at 298 K. in the flame, and E is the efficiency

of atomization. The factors Q, 0, T, and e can be measured

experimentally. The parameters nT and nh98 can be taken from

the tables given in the thesis by Zaer (71). For the flames

used in this study nT and n298 were approximately equal.











TABLE 8


CONVERSION OF SOLUTION CONCENTRATION
IN MOLES/L. TO ATOMS/CM.3
IN FLAME GASES FOR OXYGEN-HYDROGEN FLAMES


S02 ) 02 ) H20 T,O K
V=2__


2000
2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000


0.5 0.2 2660


0.5

0.5

0.5
0.4

0.4

0.4

0.4

0.5

0.5

0.5

0.5

0.6

0.6

0.6

0.6

0.5

0.5

0.5


0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5
1.0

2.0

0.2

0.5
1.0

2.0

0.2

0.5

1.0


2588

2520

2450

2695

2636

2585

2545

2650

2600

2550

2510

2607

2567

2514

2452

2667

2618

2574


3.46 x 1015
8.65 x 1015

1.77 x 1016

3.64 x 1016

2.25 x 1015

5.43 x 1015

1.11 x 1016

2.25 x 1016

2.70 x 1015

6.86 x 1015

1.40 x 1016

2.85 x 1016

3.0 x 1015

7.6 x 1015

1.55 x 1016

3.18 x 1016

2.31 x 1015

5.76 x 1015

1.16 x 1016

2.34 x 1016


Effi-
ciency

0.92

0.80

0.68

0.51

0.96

0.78

0.63

0.49

0.97
0.88

0.75

0.54

0.91

0.67

0.59

0.41

0.92

0.75

0.59


3000 0.5 2.0 2542


9.84 x 1015


3.18 x 1015

6.92 x 1015

1.20 x 1016

1.85 x 1016

2.16 x 1015

4.24 x 1015

7.00 x 1015

1.10 x 1016

2.62 x 1015

6.04 x 1015

1.05 x 1016

1.54 x 1016

2.73 x 1015

5.09 x 1015

9.15 x 1015

1.30 x 1016

2.12 x 1015

4.32 x 1015

6.85 x 1015


0.42











TABLE 9


CONVERSION OF SOLUTION CONCENTRATION
IN MOLES/L. TO ATOMS/M.53
IN FLAME GASES FOR OXYGEN-ACETYLENE FLAMES


02 o _
q C2H2

3500 2.5

3500 2.5

3500 2.5

3500 2.5
4000 2.0

4000 2.0

4000 2.0

4000 2.0

4000 2.5

4000 2.5

4000 2.5

4000 2.5

4000 3.0

4000 3.0

4000 3.0

4000 3.0

4500 2.5

4500 2.5

4500 2.5

4500 2.5


H20


0.2

0.5
1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5
1.0

2.0

0.2

0.5
1.0


T. K


2842

2792

2730

2668

2904

2844

2792

2756

2906

2847

2782

2730

2920

2813

2704

2636

2946

2872

2788


3.00 x 1015

7.62 x 1015

1.56 x 1016

3.19 x 1016

2.37-x 1015

6.03 x 1015

1.23 x 1016

2.70 x 1016

2.59 x 1015

6.53 x 1015

1.34 x 1016
2.72 x 1016

2.62 x 1015

6.82 x 1015

1.41 x 1016

2.91 x 1016

2.25 x 1015

5.56 x 1015
1.19 x 1016


2.0 2700 2.46 x 1016


Effi-
ciency

0.96

0.840

0.680

0.466

0.960

0.870

0.750
0.600

0.975

0.870

0.725

0.510

0.930

0.740

0.545

0.440

0.970

0.890

0.770

0.525


2.88 x 1015

6.40 x 1015

1.06 x 1016

1.49 x 1016

2.28 x 1015

5.24 x 1015

9.24 x 1015

1.62 x 1016

2.52 x 1015

5.68 x 1015

9.72 x 1015

1.39 x 1016
2.44 x 1015

5.04 x 1015

7.69 x 1015

1.28 x 1016

2.18 x 1015

4.95 x 1015

9.16 x 1015

1.29 x 1016










To aid in the interpretation of the data given in this

study, trwo tables (Table 8 applies to H2/02 flames and Table

9 applies to C2.H2/02 flars) are given for use in converting

from solution concentration, C, in coles per liter to N, the

number of atoms per cc. of flca gas. The conversion factor,

F', given in Tables 8 and 9, is zcltiplied by C to convert

to N. F' is given by the follcring equation:



F' = X 3 x 1021 (26)
QTaT





B. Ex:pric-ntal


The equip: nt used for emission moasurements in this

study was identical to that described in Section I of this

dissertation. The equip=mnt used for absorbance measurements

was similar to that described in Section II. Vapor lamps

(Ealing Corporation, Cambridge, Nass.) were used as spectral

line light sources for the absorbance racm'rccents.

Stock solutions containing 104 p.p.m. sodium as NaC1

and 104 p.p.m. potassicu as KC1 were prepared for this study.

From the sodium stock solution, solutions ranging in concen-

tration from 1074 p.p.m. to 104 p.p.m. sodium, differing in

concentration by a factor of 10, were prepared by successive

dilution. A similar series of solutions was prepared from

the potassium stock solution.










Relative intensity measurer3nts of the line under study

were made by scanning over the line at a speed of 2 A. per

minute and recording the =aplified output signal from the

photonultiplier tube. A lc scanning speed was used in order

to allow the recorder time to properly respond to the output

signal. A 1P28 photomultiplier tube was used to study the

Na 5890 A. line and the K 4045 A. line. A 1P22 photonultiplier

tube was used on the K 7655 A. line. All coasurecsnts vore

made in duplicate.

Absorbance casurecants were nade by scanning over the

line with the lacp on and no solution in the flana and recor-

ding the output (1), then, scanning over the line with solu-

tion in the flacn and recording the output signal (I' + Ibkg),

and, finally, measuring only the signal (Ibkg) with solution

in the flace, but with radiation from the laap blocked off

by a shutter. The absorbance was then calculated by


(I)
A = log (27)
(I Ibkg (Ibg)

As the total consumption burners operated at a tempera-

ture rcage outside of the opti um region, a special burner

was constructed which would operate at a temperature within

the optimum region. This burner consisted of a Meker burner

with a chamber-type atomizer attached. It is pictured in

Figure 11.














- Meker burner

rubber
stopper

1 1/2" o.d.
plastic
tube -


9nrn. o.d.
glass tube


drain --.


\ '- air

Bckmdn 94U20
atumiztr
solution


Figure II
Chamber-type Atomizer and Burner.


fuel-+












C. Results and Conclusions

Relative intensity data for the Na 5890 A. line for given

conditions in the H2/02 flare are given in Tables 10 through

15 and for given conditions in the C2E2/02 flame in Tables

16 through 21. Similar data are given for the K 7665 A. line

in the H2/02 fla-e in Tables 22 through 26, for the K 7665

A. line in the C2H2/02 flae3 in Tables 27 through 30, for

the K 4045 A. line in the H2/02 flaon in Tables 31 through

34, and for the K 4045 A. line in the C2E2/02 flame in Tables

35 through 38. Some data are given for the leker burner flame

with chacber-type atomizer in Table 39 for the Na 5890 A.

line and Table 40 for the K 7665 A. line.

If relative intensity data are plotted versus concentra-

tion for any given flca_ condition, curves similar to those

given in Figure 12 for a H2/02 fla=i and in Figure 13 for

an C2E2/02 flame result, i.e., a reversed S-shaped curve with
a linear center portion. The slopes of the linear center

portion are approximately the same for all conditions of the

H 2/02 flace and approdinately the same for all conditions

of the C202/02 flaca for a given spectral line.

If the concentrations at which the relative intensity

versus concentration curves deviate from linearity (at both

high and low ends) are plotted versus the temperatures of











0 02 = 4000 cc. per minute
) 02/0 C2H2 = 2.5
Medium Bore Burner
lit. = 2.0 cm.
Temperature = 2782 K.
$ H20 = 1.0 cc. per minute


Na 5890 A. Line.


I I I I


log I

Figure 12.


Example of a Working Curve for a C2H2/02 Flame.


18




16 -




14 -


O
r-


10 -











- 18 -


16 02 = 2500 cc. per minute
0 S/,H2 = 0.5
Medium Bore Burner
Ht. = 1.5 cm.
Temperature = 2550 K.
14 Q -120 = 1.0 cc. per minute



iJ 12



10 Na 5890 A. Line.
O


8 i i -
0.5 1 2 3 4 5
log I.

Figure 13.


Example of a working Curve in a H2/02 Flame.










the flames from rhich the data aas obtained, an optimum re-

gion is defined for the total consumption burner at a given

height. Such plots are given for the Na 5890 A. line at a

height of 1.5 cm. above the burner tip for the H2/02 flame

in Figure 14 and for the same line at a height of 2 cm. for

the C2H2/02 flame in Figure 15. These heights represent av-

erage heights that are used with these flames. The regions

in between the two curves in each figure represent optimam

concentration ranges for the analysis of Na in the respective

flames. The curve defining the upper limit of the optim:u

region in Figure 14 has'zero slope. This indicates that self

absorption is the primary limiting factor, as self absorption

is approximately ind encnnt of temperature. This upper limit

agrees with the upper limit predicted in Figure 8. Figure

8 also indicates that in this concentration and temperature

range, compound dissociation should not be a problem. If

compound dissociation :ere a problem, a slope greater than

zero would be expected for the upper limit curve.

The curve for the lower limit of the optimum region in

Figure 14 has a positive slope. This suggests that, in addi-

tion to the experimental detection limit, ionization is an

important factor at the lower concentration levels. The cur-

ves in Figure 8 indicate that the detection limit should not

be the primary limiting factor in this region, but rather

ionization should be the dominant factor. The calculated















0
-O 0


0
0
0 o o o

Medium Bore Burner
Ht. = 1.5 cm. above Tip


-I I p


2500


2700


2600

T, K

Figure 14.


Experimental Limits for Na 5890 A. Line in H2/02 Flames.


14 -


12 -


S10-
1 10 -


2400















no


0
0
0
----- ""'"


Medium Bore Burner
Ht. = 2.0 cm. above Tip


2900


2800

T, OK.

Figure 15.


Experimental Limits for Na 5890 A. Line in C2H2/02 Flames.


14 -


12-


10-


-I -


0 -
2600


2700


__ __


--*I *f____ ___ f .-" --- ---- /


f\










limit of detection curve represents an ideal situation. The

accuracy of the calculated limit of detection curves is li-

mited by the accuracy of the instrumental and spectral data

available and necessary for calculation of these curves.

The theoretical ionization curve is also an approximation,

as all equilibria in a flae have certain kinetic limitations

placed on them. However, the positive slope of the experi-

mental lower limit curve indicates that ionization is proba-

bly the major problem.

The curves for the Na 5890 A. line in C2H2/02 flames,

as given in Figure 15, follow similar trends. The upper li-

mit curve has zero slope and is approximately the same value

as the up;ar limit in Figure 14. The lower limit curve has

similar shape as that for the H2/02 curve, but it is slightly

higher in value. This would be expected due to higher back-

ground noise and higher temperatures of C2E2/02 flames.

Using the data obtained in this study, such curves as

those given in Figures 14 and 15 for Na would not be of as

much value for the K 7665 A. line, especially the curves

cfi nfrn the lower concentration limit. This is because the

K 7655 A. resonance line is outside of the normal operating

range of the detector used (a iP22 photomultiplier tube),

and the sensitivity is so=sehat limited. The lowest concen-

tration level detectable in the H2/02 flame was 1 p.p.m.,

while the lowest concentration level detectable in the C2H2/02









flame was 10 p.p.m. With the proper detector, detection li-

mits on the order of 10-2 p.p.m. are normally claimed (33).

The upper limit before self absorption became a problem was

approximately 102 p.p.m. in the H2/02 flame and slightly

greater than 102 p.p.m. in the C2H2/02 flame. These upper

limits are similar to those predicted by the self absorption

curves in Figure 9.

For the K 4045 A. line, the working curves are approxi-

mately linear from the limit of detection of this line, which

is slightly less than 10 p.p.m. in the H2/02 flame and slightly

greater than 10 p.p.m. in the C2H2/02 flame, to 104 p.p.m.

at the upper limit. This represents the upper limit predic-

ted by the self absorption curve in Figure 10. It is possi-

ble to use higher concentrations of potassium when using the

K 4045 A. line for analysis, rather than the resonance doub-

let, This is because of the difference in transition proba-

bilities between the two lines.

As it was impossible to obtain temperatures low enough

to be in the optimum ranges as predicted by the curves in

Figures 8, 9, and 10 with total consumption atomizer burners,

a Meker burner was used with natural gas and air for a flame

of lower temperature. The temperature of this flame, as de-

termined by the two-line method, was about 20000 K. at the

conditions at which it was operated. Using this flame, the

Na 5890 A. line was detectable at a concentration of 10"1










p.p.m., and the working curve was linear from 1 p.p.a. to

almost 103 p.p.m. This experimentally determined optimum

region corresponds to values of N from 8 x 108 to 8 x 1011

atoms per cc. of flame gas. The range as predicted by the

curves in Figure 8 was from 2.2 x 109 to 8 x 1011 atoms per

cc. of flame gas. The C 7655 A. line was detectable at a

concentration of 10 pp.p.., and the working curve was linear

irom 10 p.p.n. to almost 103 p.p.n. This corresponds to a

range of values of N from 4.6 x 109 to 4.6 x 1011 atoms per

cc. of flame gas. The theoretically predicted range from

Figure 9 was quite narrow. However, the lower limit of de-

tection was predicted at 8 x 109 atoms per cc. of flame gas,

and the optimum upper limit as governed by self absorption

was predicted to be 6.3 :: 1011.

Attclic absorption data for the Na 5890 A. line for the

H2/02 flame are given if Table 41 and for the C2E2/02 flame

in Table 42. Similar data for the K 7665 A. line are given

in Table 43 for the V2/02 flamc and in Table 44 for the C2E2/02

flace. No absorption measurements ware made with the K 40'5

A. line. As can be seen from Figure 10, this line has quite

a high limit of detection and is not suited for atomic absorp-

tion. Some cmasurermnts were made using the Meker burner

flame, and the results for the Na 5890 A. line are given in

Table 45 and for the K 76!5 A. line in Table 46.

On inspection of the data for Na and K in both types










of flac=s, it is noted that the order of absorbance for a

given concentration and solution flo:? rate for the flames

of each type is fuel rich stoichio-etric oxygen rich.

This can be partly explained on the basis of the differences

in sample introduction efficiency and tidth of flace gases

for each fla=e type. The order of efficiencies and flame

r:idths is the sae as that for the order of absorbancies.

The lowest detectable concentration of Na by atoEic ab-

sorption spectroscopy of the Na 5890 A. line in E2/02 flames

was 1 x C11 atoms per cc. of flanm gas. The lowest experi-

mentally detectable concentration of Na in an C2E2/02 flame

was also 1 x 1011 atc-:s per cc. The lo::est detectable con-

centration of Na by atoic absorption of the Na 5890 A. line

as predicted by the calculated curve in Figure 8 was 4 x 108

atoms per cc. of flaar gas. The lo:-:est liit of potassiln

expericentclly detectable by absorption of the K 7665 A. line

:as 5 x 1011 atoms per cc. for a E2/02 flane and 6 x 1011

tons per cc. for an C2E2/02 flare. The lniit predicted by

calculation was 1.2 x 1010 atoms per cc. The calculations

were based on an equation (70) hhich assnued the use of an

optiEcu slit jidth. This calculation also assumed that ther-

nal emission of the line studied rwas negligible and did not

account for flanr flicker. The more extensive equation (70)

which does not assucs an optimun slit width takes nore instru-

mental factors into consideration and would probably give

values closer to those realized experimentally.








The only factor limiting the upper concentration limit

in atomic absorption spectroscopy, other than incomplete com-

pound dissociation, is the ability to read small signals ac-

curately. This is because the intensity of the line with solu-

tion in the flames tends toward zero, after flame emission

has been corrected for, as the solution concentration increases.

The flames used in this study were far from ideal for

atomic absorption spectroscopy. The flames usually used (25)

have much longer optical path lengths and lower temperatures

than the ones used in this study. Only a flame with a tempera-

ture high enough to insure compound dissociation is needed.

The higher temperature flames have more intense backgrounds,

and the emission of the element under study becomes a problem,

especiallyin the case of sodium and potassium.

Optimum conditions for emission or absorption analysis of

Na and K introduced as NaCl and KC1 into H2/02 or C2H2/02 flames

via total consumption burners can be determined from the N-T

plots given in Figures 8, 9, and 10. First convert the two

limiting concentrations from moles per 1. to atoms per cc. as

previously described. Then using the N-T plots, draw lines

parallel to the T axis at the given N values. This defines

the limiting T values, and the proper flame types can be se-

lected accordingly. Reference can also be made to the tables

of experimental data to enable the analyst to choose conditions

giving maximum intensities for emission analysis. If a total

consumption burner is used, and it is not possible to obtain

a temperature in the optimum region, only the upper and lower

concentration limits will be important.











TABLE 10


SODIUM IN THE HYDROGEN FLAME AT 5890 A.


(Solution

) H20


2000

2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000


Concentration =,

NT'


) 02


02

0.5
0.5

0.5

0.5

0.4

0.4

0.4

0.4

0.5

0.5

0.5

0.5

0.6

0.6

0.6

0.6

0.5

0.5

0.5


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


10-2 ppm.)

Log NT


x 109

x 109

x 109

x 109

x 108

x 109

x 109
x 109

x 109

x 109

x 109

x 109

x 109

x 109

x 109

x 109
x 108
x 109

x 109


I Log I


1.39

3.01

5.24

8.06

9.40

1.84

3.06

4.80

1.135

2.62

4.35

6.69

1.18

2.21

3.97

5.66

9.2

1.88

2.98


9.143

9.479

9.720

9.906

8.973

9.265

9.483

9.681

9.055

9.418

9.656

9.826

9.072

9.344

9.599

9.753

8.964

9.274

9.474

9.631


-

0.954

1.114

1.042

1.362

0.903

0.954

1.042

0.845

0.699

0.778

0.845

0

0.954

0.954

0.778

0.602



0.699
0.699


4.27 .x 109


3000


6 0.778


0.5











TABLE 11


SODIUM IN THE HYDROGEN FLAME AT 5890 A.


(Solution

H20


2000

2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000


Concentration =

NT'


S02


S02

0.5
0.5

0.5

0.5

0.4

0.4

0.4

0.4

0.5

0.5

0.5

0.5

0.6

0.6

0.6

0.6
0.6

0.5

0.5

0.5


10-1 ppm.)

Log NT'


x 1010

x 1010

x 1010

x 1010

x-109

x 1010

x 1010

x 1010

x 1010

x 1010
x 1010

x 1010

x 1010

x 1010

x 1010

x 1010

x 109

x 1010

x 1010


I Log I


10.143

10.479

10.720

10.906

9.973

10.265

10.483

10.681

10.055

10.418

10.656

10.826

10.072

10.344

10.599

10.753

9.964

10.274

10.474


62

65

78

88

74

173

219

289

36

85

115

141

38

55

63

69

27

81

126


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


1.782

1.813

1.892

1.945

1.869

2.238

2.340

2.461

1.556

1.930

2.061

2.149

1.580

1.740

1.799

1.839

1.432

1.909

2.100


4.27 x 1010 10.631


1.39

3.01

5.24

8.06

9.40

1.84

3.06

4.80

1.135

2.62

4.35

6.69

1.18

2.21

3.97

5.66

9.2

1.88

2.98


3000


0.5


176 2.246











TABLE 12


SODIUM IN THE HYDROGEN FLAME AT 5890 A.


(Solution Concentration =


SH20


NT'


2000

2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000

3000


b 02


( 02

0.5
0.5

0.5

0.5

0.4

0.4

0.4

0.4

0.5

0.5

0.5

0.5

0.6

0.6

0.6

0.6

0.5

0.5

0.5

0.5
0.5


10 ppm.)

Log NT'


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


I Log I


1.39

3.01

5.24

4.06

9.40

1.84

3.06

4.80

1.135

2.62

4.35

6.69

1.18

2.21

3.97

5.66

9.2

1.88

2.98

4.27


x 1012

x 1012

x 1012

x 101
12


x 1011

x 1012

x 1012

x 1012

x 102

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012
x 1012


12.143

12.479

12.720

12.906

11.973

12.265

12.483

12.681

12.055

12.418

12.656

12.826

12.072

12.344

12.599

12.753

11.964

12.274

12.474

12.631


2776

5570

6568

6670

6071

14463

18563

23566

2981

7780

10582

11784

1690

3992

4492

4792

3669

6872

10676

13380


3.443

3.746

3.818

3.824

3.783

4.160

4.268

4.372

3.475

3.891

4.024

4.074

3.228

3.600

3.653

3.680

3.564

3.837

4.028

4.126










TABLE 13


SODIUM IN THE HYDROGEN FLAME AT 5890 A.
(Solution Concentration = 1 ppm.)

0T2 0 H20o NT' Log NT' I Log I
i42
2000 0.5 0.2 1.39 x 1011 11.143 726 2.861
2000 0.5 0.5 3.01 x 1011 11.479 1080 3.033
2000 0.5 1.0 5.24 x 1011 11.720 1698 3.230
2000 0.5 2.0 8.06 x 1011 11.906 2120 3.326
2500 0.4 0.2 9.40 *x 1010 10.973 871 2.940
2500 0.4 0.5 1.84 x 1011 11.265 1728 3,238

2500 0.4 1.0 3.06 x 1011 11.483 2793 3.446
2500 0.4 2.0 4.80 x o101 11.681 3826 3.583
2500 0.5 0.2 1.135 x 1011 11.055 781 2.893
2500 0.5 0.5 2.62 x 1011 11.418 1310 3.118
2500 0.5 1.0 4.35 x 1011 11.656 2182 3.338
2500 0.5 2.0 6.69 x 1011 11.826 2714 3.434
2500 0.6 0.2 1.18 x 1011 11.072 470 2.672
2500 0.6 0.5 2.21 x 1011 11.344 1052 3.022
2500 0.6 1.0 3.97 x 1011 11.599 1342 3.128
2500 0.6 2.0 5.66 x 1011 11.753 1572 3.197
3000 0.5 0.2 9.2 x 1010 10.964 448 2.652
3000 0.5 0.5 1.88 x 1011 11.274 885 2.947
3000 0.5 1.0 2.98 x 1011 11.474 1684 3.227
3000 0.5 2.0 4.27 x 1011 11.631 2351 3.371











TABLE 14


SODIUM IN THE HYDROGEN FLAME AT 5890 A.


(Solution Concentration =


102 ppm.)


I Log I


S02

2000

2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000

3000


0.5

0.5

0.5

0.5

0.4

0.4

0.4

0.4

0.5

0.5

0.5

0.5

0.6

0.6

0.6

0.6

0.5

0.5

0.5

0.5


SH20O


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


1.39

3.01

5.24

4.06

9.40

1.84

3.06

4.80

1.135
2.62

4.35

6.69

1.18

2.21

3.97

5.66

9.2

1.88

2.98

4.27


x 1013

x 1013

x 1013

x 1013

x 1012

x 1013

x 1013

x 1013

x 1013

x 1013

x 1013

x 1013
x 1013
x 1013

x 1013
x 1013

x 1012

x 1013

x 1013

x 1013


NT'


Log NT'


13.143

13.479

13.720

13.906

12.973

13.265

13.483

13.681

13.055

13.418

13.656

13.826

13.072

13.344

13.599

13.753

12.964

13.274

13.474

13.631


26276

45270

47968

42970

36971

78963

90963

83966

23981

46980

54982

47984

12990

26992

32992

29992

24969

50972

57976

56980


4.420

4.656

4.681

4.633

4.567

4.897

4.959

4.924

4.379

4.680

4.739

4.680

4.113

4.430

4.518

4.462

4.397

4.707

4.763

4.755











TABLE 15


SODIUM IN THE HYDROGEN FLAME AT 5890 A.


(Solution Concentration


( H20


( 02


NT'


02

0.5

0.5

0.5

0.5

0.4

0,4
0.4

0.4
0.54
0.5

0.5

0.5

0.6

0.6

0.6
0.6

0.6

0.5

0.5

0.5


= 103 ppm.

Log NT'


x 104

x 1014

x 1014

x 1014
x 1014

x 1014

x 1013
x 1014

x 1014



x 1014
x 1014



x 1014
x 104

x 1014




x 1014
x 101
x 1o14




x 1014


x 1014
x 101
x 1014


I Log I


14.143

14.479

14.720

14.906

13.973

14.265

14.483

14.681

14.055

14.418

14.656

14.826

14.072

14.344

14.599

14.753

13.964

14.274

14.474


60976

76970

69968

50970

119971

139963

135963

102966

71981

84980

82982

63984

45990

51992

52992

41992

85969

92972

88976


2.0 4.27 x 1014 14.631


1.39

3.01

5.24

4.06

9.40

1.84

3.06

4.80

1.135

2.62

4.35

6.69

1.18

2.21

3.97

5.66

9.2

1.88

2.98


2000

2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0


3000 0.5


4.785

4.886

4.844

4.707

5.079

5.146

5.133

5.012

4.857

4.928

4.918

4.805

4.662

4.716

4.724

4.623

4.934

4.968

4.949


71980 4.856











TABLE 16


SODIUM IN THE ACETYLENE FLAME AT 5890 A.


(Solution Concentration =


10-2 ppm.)


S02 c 02

3500 2.5

3500 2.5

3500 2.5

3500 2.5

4000 2.0

4000 2.0

4000 2.0

4000 2.0

4000 2.5

4000 2.5

4000 2.5

4000 2.5

4000 3.0

4000 3.0

4000 3.0

4000 3.0

4500 2.5

4500 2.5

4500 2.5


4500


2.5


NT'


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


1.25 x

2.79 x

4.62 x

6.49 x

9.96 x

2.28 x

4.01 x

7.32 x

1.10 x

2.47 x

4.22 x

6.02 x

1.06 x

2.19 x

3.34 x

5.36 x

9.49 x

2.16 x

3.98 x


Log NT'


109

109

109

109

108

109

109

109

109

109

109

Io09

109

109

109

109

108


109
109


5.62 x 109


9.097

9.446

9.665

9.812

8.996

9.358

9.604

9.865

9.042

9.393

9.625

9.780

9.025

9.340

9.524

9.729

8.977

9.335

9.600

9.750


I Log I


28

66

75

140

173

192

420

525

34

37

102

86


16

43

57

89

91

131


1.447

1.820

1.875

2.146

2.238

2.283

2.624

2.720

1.532

1.568

2.009

1.935


1 .204

1.634

1.756

1.949

1.959

2.118


185 2.267


H20











TABLE 17


SODIUM IN THE ACETYLENE FLAME AT 5890 A.


(Solution


4 02 A 02
3 2-0 2
3500 2.5

3500 2.5

3500 2.5

3500 2.5

4000 2.0

4000 2.0

4000 2.0

4000 2.0

4000 2.5

4000 2.5

4000 2.5

4000 2.5

4000 3.0

4000 3.0

4000 3.0

400 3.0

4500 2.5

4500 2.5

4500 2.5

4500 2.5


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


Concentration

NT'


1.25

2.79

4.62

6.49

9.90

2.28

4.01

7.32

1.10

2.47

4.22

6.02

1.06

2.19

3.24

5.36

9.49

2.16

3.98

5.62


1010

1010

1010

1010

109

1010

1010

1010

1010

1010

1010
1010

1010

1010

1010

1010

109
1010

1010

1010


= 10-1 ppm.)


Log NT'


10.097

10.446

10.665

10.812

9.996

10.358

10.604

10.865

10.042

10.393

10.625

10.780

10.025

10.340

10.524

10.729

9.977

10.335

10.600

10.750


I Log I


126

123

273

413

280

546

767

1133

133
228

400

684

56

143

204

269

159
248

295

355


2.100

2.090

2.436

2.616

2.447

2.738

2.885

3.054

2.124

2.358

2.602

2.835

1.748

2.155

2.310

2.430

2.202

2.394

2.470

2.550


H20












TABLE 18


SODIUM IN THE ACETYLENE FLAME AT 5890 A.


0) 2 02

3500 2.5
3500 2.5

3500 2.5

3500 2.5

4000 2.0

4000 2.0

4000 2.0

4000 2.0

4000 2.5

4000 2.5

4000 2.5

4000 2.5

4000 3.0

4000 3.0

4000 3.0

4000 3.0
4000 3.0

4500 2.5

4500 2.5

4500 2.5


4500


2.5


(Solution

(H20


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


Concentration

NT'


1.25

2.79

4.62

6.49

9.90

2.28

4.01

7.32

1.10

2.47

4.22

6.02

1.06

2.19

3.24

5.36

9.49

2.16

3.98


x 1011

x 1011

x 1011

x 1011

x 1010

x 1011

x 1011

x 1011

x 1011

x 1011

x 1011

x 1011

x 1011

x 1011

x 1011
x 1011

x 1010

x 1011

x 1011


5.62 x 1011


= 1 ppm.)

Log NT'


11.097

11.446

11.665

11.812

10.996

11.358

11.604

11.865

11.042

11.393
11.625

11.780

11.025

11.340

11.524

11.729

10.977

11.335

11.600

11.750


I Log I


750

2160

2833

3953

1990

5166

7717

12263

1180

2669

3914

5634

456

1213

1714

2339

969

2618

3275


2.875

3.335

3.452

3.596

3.299

3.753

3.888

4.089

3.072

3.427

3.592

3.751
2.660

3.084

3.234

3.369

2.986

3.418

3.515


5592 3.747











TABLE 19


SODIUM IN THE ACETYLENE FLAME AT 5890 A.


(Solution


4 H20


Concentration

NT'


= 10 ppm.)

Log NT'


I Log I


02 A002

3500 2.5

3500 2.5

3500 2.5

3500 2.5

4000 2.0

4000 2.0

4000 2.0

4000 2.0

4000 2.5

4000 2.5

4000 2.5

4000 2.5

4000 3.0

4000 3.0

4000 3.0

4000 3.0

4500 2.5

4500 2.5

4500 2.5

4500 2.5


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


1.25

2.79

4.62

6.49

9.90
2.28

4.01

7.32

1.10

2.47

4.22

6.02

1.06

2.19

3.24

5.36

9.49

2.16

3.98

5.62


x 1012

x 1012

x 1012

x 1012

x 1011

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012

x 1012
x 1012

x 1012

x 1012

x 1012

x 1012


x 1012
x 1012


12.097

12.446

12.665

12.812

11.996

12.358

12.604

12.865

12.042

12.393

12.625

12.780

12.025

12.340

12.524

12.729

11.977

12.335

*12.600

12.750


10600

20200

26173

24153

15930

40936

60063

89863

11920

26209

36914

50234

5246

13353

17664

22979

7629

18588

25865

35432


4.025

4.305

4.418

4.383

4.202

4.612

4.779

4.953

4.076

4.419

4.567

4.700

3.720

4.126

4,247

4.361

3.882

4.269

4.413

4.550











TABLE 20


SODIUM IN THE A.ETYLE F ..J- AT 5890 A.


S02 02

3500 2.5

3500 2.5

3500 2.5

3500 2.5

4000 2.0

4000 2.0

4000 2.0

4000 2.0

4000 2.5

4000 2.5

4000 2.5

4000 2.5

4000 3.0

4000 3.0

4000 3.0

4000 3.0

4500 2.5

4500 2.5

4500 2.5


(Solution


( H20


0.2

0.5
1.0

2.0

0.2

0.5

1.0

2.0

0.2

0,,5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


Concentration

NIn


1.25

2.79

4.62

6.49

9.90
2.28

4. 01

7.32

1 ,10

2.47

4.22

6.02
I.o6
1.06

2.19

.34

5.36

9.49
2.16

3.98


1013

1013

1013
1013
1012

1013

1053


1013

1013
1013

1013
1013

1013

1013

1013
1013

1013
1013


= 102 ppm.)


Log NT'

13.097

13.446

13.665

13.812

12.996

13.358

13.604


13.042






13.025
13,780


13.343

13.524

13.729

12.977

13.335

13.600


I Log I


71900

100900

108873

104853

109930

215936

249930

239930

81920

3!9094

1 ) ,4






85964.


58?6

10738S

125865


13.750 126832 5.103


4.855

5.004

5.036

5.020

5.041

5.334

5.398

5.380

4.913

5.146

5.178

5.161

4.629

4.892

4.934

4.914

. 770


5,. 10
5.100


4500


2.5


5.62 x 1013










TABLE 21


SODIUM IN THE ACETYLENE FLAME AT 5390 A.
(Solution Concentration = 103 ppm.)
( 02 ) 02 H20 NTr Log NT' I Log I

3500 2.5 0.2 1.25 x 1014 14.097 180900 5.257
3500 2.5 0.5 2.79 x 1014 14.446 193900 5.288
3500 2.5 1.0 4.62 x 1014 14.665 174873 5.242
3500 2.5 2.0 6.49 x 1014 14.812 140853 5.149
4000 2.0 0.2 9.90 x 1013 13.996 352930 5.548
4000 2.0 0.5 2.28 x 1014 14.355 385936 5.587
4000 2.0 1.0 4.01 x 1014 14.604 365917 5.564
4000 2.0 2.0 7.32 x 1014 14.865 -02,66 5.482
4000 2.5 0.2 1.01 x 1014 14.042 222-20 5.348
4000 2.5 0.5 2.47 x 1014 14.393 255C09 5.408
4000 2.5 1.0 4.22 x 1014 14.625 23904 5.378
4000 2.5 2.0 6.02 x 1014 14.780 196.34 5.294
4000 3.0 0.2 1.06 x 1014 14.025 115946 5.364
4000 3.0 0.5 2.19 x 1014 14.340 131953 5.120
4000 3.0 1.0 3.34 x 1014 14.524 118964 5.076
4000 3.0 2.0 5.36 x 1014 14.729 95969 4.982
4500 2.5 0.2 9.49 x 1013 13.977 146969 5.167
4500 2.5 0.5 2.16 x 1014 14.335 182888 5.262
4500 2.5 1.0 3.98 x 1014 14.600 182865 5.262
4500 2.5 2.0 5.62 x 1014 14.750 159832 5.204











TABLE 22


POTASSIUM IN THE HYDROGEN FLAME AT 7665 A.


02 O2

2000 0.5
2000 0.5


2000 0.5
2000 0.5

2500 0.4

2500 0.4

2500 0.4

2500 0.4

2500 0.5

2500 0.5

2500 0.5

2500 0.5

2500 0.6

2500 0.6

2500 0.6
2500 0.6

2500 0.6

3000 0.5

3000 0.5

3000 0.5


3000


0.5


(Solution

SH20O


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


Concentration =

NT'


8.11

1.76

3.06

4.72

5.50

1.08

1.79

2.71

6.68

1.54

2.68

3.93

6.96

1.30

2.33

3.32

5.41

1.10

1.75


x 1010

x 1011

x 1011

x 1011

x 1010

x 1011

x 101

x 1011

x 1010
x 101



x 10

x 1011
x 10


x 101
x 1010

x 1011

x 1011

x 1011

x 1010
x 1011






x 1011
x 101


2.51 x 1011


1 ppm.)

Log NT'


10.909

11.246

11.486

11.674

10.741

11.034

11.253

11.433

10.824

1 .1 88

11.428

11.594

10.843

11.114

11.367

11.521

10.734

11.042

11.243


I Log I


1.5

3

3.5

1


i.

6

I





5
1.5

3

3.5


1

2

3


1

3


11.400 4


-

0.176

0.477

0.544

0

0.398

0.602

0.778



0.176

0.477

0.544


0

0.301

0.477


0

0.477

0.602











TABLE 23


POTASSIUM IN THE HYDROGEN FLAME AT 7665 A.


(Solution

H20


2000

2000

2000

2000

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

2500

3000

3000

3000

3000


Concentration =

NT'


02


0 op

0.5
0.5

0.5

0.5

0.4

0.4

0.4

0.4

0.5

0.5

0.5

0.5

0.6

0.6

0.6
0.6


0.5

0.5

0.5

0.5


10 ppm.)

Log NT'


I Log I


0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0

0.2

0.5

1.0

2.0


8.11

1.76

3.01

4.72

5.50

1.08

1.79

2.71

6.68

1.54

2.68

3.93

6.96

1.30

2.33

3.32

5.41

1.01

1.75

2.51


1011

1012

1012

1012

1011

1012

1012

1012

1011

1012

1012

1012

1012

1012

1012

1011

1012
1012

1012


11.909

12.246

12.486

12.674

11.741

12.034

12.253

12.433

11.824

12.188

12.428

12.594

11.843

12.114

12.367

12.521

11.734

12.042

12.243

12.400


0.602

1.079

1.279

1.398

0.954

1.415

1.653

1.806

0.778

S.146

1.398

1.505

0.602

1.000

1.176

1.230

0.903

1.342

1.532

1.672




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs