REPORT NUMBER 1

                       August 7n 1963
                     Raymond 9» Rawson
                      James Rn Dillon
     Flame photometric determinations for calcium and strontium
a^e an established portion of the routine work of this
laboratory°  The use of flame photometry provides the most
rapid and economical method of performing these determinations-;
but Is subject to extensive interference problems»  The
inherent instability of the instrumental system requires
considerable operating skill to obtain acceptable results»
     An investigation -was made of a reversed-gas flow system
(to the atomiser-burner) to determine if improved operating
characteristics could be achieved in this manner*  The results
show an increased flame temperature /> without heating of the
atcsaiser-burner during operation, and an impressive increase
in flame stability resulting from elimination of the deposition
of combustiora products on the burner tip,.  The band spread
can be reduced by the increase in sensitivity which reduces
the interference problems <>
     By comparison with conventional burner operations the
reversed gas flow system results in markedly increased accuracy
and precision«  Establishment of routine procedures using this
system is advised *

-------
     Certain elements introduced into a high-temperature
flame will emit a characteristic visible spectrum <>  In.  flame
photometric analysis t these characteristic radiations are
measured and compared to a standard to provide a quantitative
estimate of specific elements c,
     In the normal operation of the Beckman model DO flams
photometer a the sample is introduced into ara oxyaeetyiene
flame with oxygen used as the spraying gas  Cfig» IK  The
flame is formed at the capillary tip, resulting ±n a heating
of the burner and consequent carbon and salt enerwstment of
the capillary tip«  These deposits are picked up in the flame-.
teausing fluctuations in the flame temperature and position,,
resulting in an unstable system *  The effect of this on normal
operation is to cause flame photometry to be largely an
empirical procedure compounded by inherent instability and
short operating intervals with extensive interference problems >.
Nevertheless,? flame photometry has proven to be a most
reliable and time saving procedure for the determination of
eaiciusst strontium * sodium „ potassium and lithium on a
routine basis o
                                                \
     A publication by Loken* Tealc and Eisenbery  proposed
a procedure for calcium using a reversed oxyaeetylene
The papsr indicated that by modifying the standard Bsckican
BU flame photometer to use acetylene as the spraying gas.-,
the problems inherent ±n normal operation are largely
that isj a higher degree of stability and repeatability of
the flam© photometric system is achieved °  It is the
purpose of this investigation to determine if this modification
would result in improvement of the flame photometric psceednares
that this laboratory has in use for calcium and strontium
analysis o

-------
      The following apparatus was used  for  th:i.s i'3W3£sii£?-is';:ta;!,s
                  an model BUJ Spectrophotosietei?
                     $9200 fflame Attaelment
            JSsekfitan $4030 Medium Bore Qxygexa^aewtylssKs l-fey ne"'
      This instrument was operated by the coffi^eatic^.sil v.!araual
TQ'i^'CjS'G* C»ii*j^. (^t ci«3 jQ^ijj^.  T£JQiG2 i^icUlitil^j.5l^2X»v\2>^2«£ £5 ST^S^O'jTtjj^^??5^0i(Cj •:* j'-^vJ^ri^  'vv--• ^ '.i..
the  except ion that the gas  lines inside the saaaciaete:^ ho?s8;h!g  »
were altered a as  per fig0 2j to use acetylene as the sp::&y ? r.g
      *T-h© gauges  on the ojeygen and acetylene fesuiiks CI.T© s;e-«- EU;
IS pcScio aad 12 p0Soio respectively»  She ae©l;yl,o,,io and the closed aeedle ^-alve is opened a?3©ut  ©P® ha.'.l.f
tuarao  ^ie flar.e is lite   If it does not stay litf.  the a ths
oxygen is turned down.? with  special care not. fco get it 'ceo
r.o;-y  Jindicated by a very bright,- white flaasejo  Thco SJ
acjTastaisKt of gas pressures  is then made o  'Sfte bl^.e ce:fi
cone at the fiarae base should be frosa 10 to 25 :ata high..
v/ater is sprayed through the burner :f the center eo.v:.G
decrease to 1 ma or less°  Five mm appears to &a ar
height.   ?he larger ths eenter cone,-, the large:? ths e,s
of bae2r.grouKid! light..
      cJ?Le Mirror  is thaB adjusted to gi^e a Eie.3g;;.3.Mffi  raj
       a sasnple  is  beiag sprayed o   This  is dose with  C
                                 2
         or strontiiam standard 0
  re  saade u^p iss  the following eoncentratioBS.  0<, 001  g/Jl -
    5 g/l(- OoOiO g/lff 0,015 g/1., OcC20 g/1,  amd Od g/;'_-
      arcl (Serves  "were thera iznn and interfesesiGSf? ccaekest,.
      After final adjustment is made., the  oxygen pj-3ssu.re is
      aisLsd at 1<>2 poSoi» and  the acetylene pressure st 7.C p. B

-------
^he spscrt&'al  lines at 422.7 m\.\ ©re  xased for Sa; 460:-7 :;e;-  •'.;?:
i>SS>o3  ror  Kfeu  and 766»5 for K^   In  ea
     •Ehe  following changes in operating characteristics of
til© BscCor.an D?J flame photometer were  observed as a result of
reversing tae cssygen arod acetylene  feed to the atcBizeK'-buza
     Io   ^he oxygen pressiase is redweed f^csa ojs®~hajbc to
two»thi:rcs of that required by conventional opsration<-  uvhaa
results in a decreased flow rate  of the sample solv-t.tom a:cc!
less cf a eool3ing effect in tine flazaec   Aa appssesifcly higher
             flaBRS results whieh increases the Isxstr
ssnsitl'.?ity ®jid introdwees a background inter f are sase
     ic'i C_ bands "* in the visible and  near^wltra violet
         <£>
        -.  This interference ean be eliminated b
         srrt  of the foswsing sai^2?©r   ffig= 3}^
fcetssed .• the slit width cam be reduced by a fasten cf tes;.  r.
a result of the increase in sensitivity over th© • convert. »
                1 and 2 deaonstrat© the
                          3
                    ia>annaa3.  states  that a mean deviation,  sf
v 0<-3 transraittance seal® divisions  is acceptable,,  fhe
reversed  flams therefore has an excellent

-------
with a mean deviation of jr CU09 transmittance scale divisions»
Hot only is the reproducibility good on a short term basis,,
there is a low mean deviation of ± Oo28 mg on strontium
standards run on 10 different days..  This is compared with
a mean deviation of + 2*23 mg Sr with the conventional flame°
Because of the slower aspiration rate* standards last much
longer»
COHCLUSIOBT
     The comparative study of conventional and reversed
oxyacetylene flame photometry indicates that the reversed
system yields a greater sensitivity with reduced background
interference, a higher order of precision, and longer periods
of operationo  The combined effect is to provide a greater
confidence in the results and an increase in productive
capacity for the flame photometer without adding to the cost
of the equipment^  Routine procedures for calcium and strontium
are presently being modified to take advantage of this
modification o
REFERENCES
     • (1)  Loken^ Ho F^a Teal, J» S=, Eisenberg, E., Anal. Cjhemc
          35, 875  (1963)=
      (2) &  (3)  Beckman Instruction Manual No* 334-A, 8-9  (1957)
      (4)  Handbook of Analytical Chemistry 6-221  (1963)»
      (5)  Broida, H» P» and Heath*. Do F» J* Chemo PhysP. 26,
          223  (1962)..
     Data by R= Rawson, Jo Dillon
     Participants E» Pyzel^ E. Cooley, B, Fumagalli

-------
                                                           Acetylene
                                                    B
       Figure I.  Schematic Draining  of  oxyacetylene
                  Flames in Bectonan  oxyaeetylene burner
          &.,  With acetylene as  spraying gas - Reversed
          B»  With oxygen as spraying gas  -  Conventional
   Supplies
                                                   Hesdle Valve
Figure 2=  Arrangement of connections in Beckiaan mancajster
           housing for reversed system

-------
       0.01 g/I

            HOT slit
                                          CcCI  g/1

                                         1 0 = 05  am slit.
                                                         « peak
                                                         ji >A

                 .
            Wavelength
       Calcium -T   Off Center FOCU.S
                 --
                                        &
                                        30^^^.

                                         ''" '* Wa -/e le .ng^E
                                                   Strontium
                         Go 03 ram slit
              11'
                Ii
             ','
      Wavelength




  'r'  OoOl  g/1

      o.os
                                                              0.01  g/1
                                                                           |
                                                              0..v.5  mm slii
                                                                  Son
                                           Wavelength
o
S-"
-p
                     «      ,
                     Ca pealc
                    I
                                       ^51
& '' I

r'
rtf .':
•4J
                                      ra

                                      R5
                                      M ?i
                                                             0.03.  g/1
                                                  A
                                                  1  t
                                                     r pea
E-i
ft /C-l
       .._L.JL-.
                                                                   Is'Iarae
    _„.„,_._	.V  V

                                           Wavslength
                               Correct Focus

-------
  Oo025
14
Q)
  0*015-
5-!

O.

CO
  0 = 005
3
  OoOOl
Wavelength-422 * 7
Slit       -0*08
.Oxygen     -1  p»s-i°
Fuel       -7  p»s-i«
Spread  $0.010-0., 025)  = 27*6
                                         &
                                         4J
                                         a
                                         IO.OIGJ
              LtTTB  "29  25  JO   35

                 % Transmission
                                                                   Wavelength-460.7
                                                                   Slit      -0,08
                                                                   Oxygen    -1 p.Soi.
                                                                   Fuel      -7 p.s.i.
                                                                   Spread  (0.005~0.010)=8.2
                                                               5'0      60      70
                                                                     % Transmission
                                                                               80
    Calcium Calibration Curve  (0=001-0.025  g/1)    Strontium Calibration Curve  (0*005-0.020 g/1
    025"
•P
•
-------
TABLE
feiapis  nrui  3.0 vrlTxss  i:.~ One
nPv* "? Ti*^ CJ'Tf^ T \m "2™ J^l TiP1* <& ™?
J. Jf, f * ; Jl d g>d il! iC-i W t*Vi — J>^< Vrf CJ
Reading
67.1
67.1
67.0
S7d
66.9
66.9
66 = 9
67.3
67.1
67ol
Average 67.1
TABLE
rag Sr
6~14~63 79.5
6-17-63 7S.O
6-18-63 80.0
6-19-63 80.0
6-19-63 79 o 5
6«o *? C\ «M ft ^ £$ O O
*6* V/ W *? w W o V/
6»24-63 80 oO
6-25-63 80.0
6-25-63 80.0
6-25-63 80.0
Average 79.8
caie Deviation fe
__™«.^_^^i^f i'Hl „
0.0
0.0
0«I
0.2
0.2
Sum 0.9
12= 1 Sample Run 10
•Dill

0 9
Maan Deviation™?-;;— = 0.09
1U
0.09 Transmit tance Scale Division

Conditions •» Sr Carrier
460.7 RIM. - Wavelength
0=OS Him « Slit Width
Masciiiiisro •*> Sensitivitv
34o."2% T - Spread Between
0.015 and Co 020
1-2 p«s.io •• Oxygen
7.0 p.s*i« - Acetylene
Different Days
Deviation from
Average
0.3
0.8
0.2
0.2
0.3
0.2
0.2
0.2
0.2
0.2
Sum 2.8




Mean Deviation = T7r~ - 0.28 jag Sr





Hotes It was necessary to find the mean deviation in
rag Sr rather than % transmittance because of
the different set of conditions each day.

-------