The spectrographic determination of total barium in bone ash

ANALYTICA

THE

SPECTROGRAPHIC Xl.S. \V.

Afomic

Energy

Research

CHIMICA

ACTA

DETERMINATION BONE ASH WEBU

Establishment,

349

OF TOTAL

BARIUM

IN

ASD hi. L. \VORDINGI-IAM Wooiwich Orctsfatiox, (Cvcal Britain)

(Iicccivcd

August

Joth,

C.37

Royal

.4m_wal.

Lowlou

19G:)

ISTItODUCTION

In recent years the physiological effects produced by radioactive substances have necessitated an intensified study of the metabolism of a wide range of elements in man. The burdens of strontium* and radium” carried by human tissues have been of special interest because of the hazard to health occasioned by the long-livccl isotopes and daughters of these elements. Whilst the radiological danger to health involved in the assimilation of trace amounts of barium is probably small, the close similarit> of this clement with both strontium anti radium made it clcsirnl~lc to study its distribution in the body. The spectrographic method outlined below was clcvclopccl in response to a demand for a rapid method fog the determination of barium in bone ash.

A spectrographic method had already been clevelopecl for the detcrminntion of total strontium in bonc3 and the possibilitv of its use for the determination of barium wits investigated. In this method the sample of bone ash was ggouncl wit11 nnhyclrous copper sulphatc and graphite and pellets of the mixture burnt in a d.c. arc. It was found that this technique was not applicable to the determination of barium because the most sensitive barium lines were overlain with a band system and in any cast the sensitivity was not sufficient to cover the desired concentration rang2 of o-25 p.p.m. Espcrimcntal work was directed therefore towards developing a m&hod with improved sensitivity and in particular to the reduction of the intensity of the intcrfering band system associated with the most sensitive barium lint at 4554.0 A. . ESI’ERIM@NTAI.

Initial attempts were made to improve sensitivity by using alkali fluorides as fluxes in the hope that barium would distii selectively. This approach was not successful. The cathode layer tcchn ique4 anti intermittent a.c. arc m&hods applied clircctly to bone ash, showed no impr,ovement. An attempt was then made to obtain adequate sensitivity by using copper sulphate and graphite as in the method for the determination of strontium” but to reduce the ratio of spectrographic buffer to sample and to increase the weight of the pellet burnt in the arc to 30 mg. This method gave improved, although inadequate, sensitivity for barium and it was still necessary to suppress C.

Anal.

Chim.

Acta,

26 (rgbz)

349-354

350

31.S. W.

WEBB,

&I. I,.

WORDINGHAM

the interfering hand system. An atmosphere of argon and oxygen was used for this purpose and in addition the argon bad the effect of increasing the intensity of the barium spark lines to an adequate value; the spectrographic arc was, however, somewhat unstable in atmospheres of oxygen and argon mixed in various proportions and inadequate precision resulted. Efforts were then made to “screw-stabilise” the arc by using a “Stallwood” jet (Fig. I) to cause the mantle of misccl gases surrounding the arc plasma to rotate. With a 25/75 oxygen/argon mixture flowing at 4 l/min, optimum conditions for sensitivity and precision were obtained.

Argon/ oxygen

Fig.

I,

jet.

Stallworxl

0a LO

20

30

‘I_1 40

50

4558.5A

60

70

Tlme(sec) Fig.

2.

Enlisuicw

curvcn.

No suitable copper lines could bc found which were homologous with the barium lines and it was evident that another internal standard would have to be selected. Of the various elements csamined for this purpose by the conventional falling plate technique, lanthanum in the form of osicle was the most suitable (Fig. z) and a concentration of o.x7O/” of lanthanum as osiclc in the pellet gave a convcnicnt line density. A standard curve was obtained from synthetic standards prepared by dry grinding and successive dilution of barium sulphatc with calcium phosphate. For initial experiments calcium phosphate which had been prepared from a small sample of pure calcium metal was used; this had been shown to contain 3 p.p,m. of barium by the method of standard addition and cathode ray microphotometryb. Subsequent in-

SPECTROGRAPHIC

DETERMIXATIOS

Ba

OF

IS

BOSE

ASH

3s

vestigations showed that the barium content of analytical grade calcium carbonate could be recluced to a satisfactory level by an ion-eschange technique. The method adopted was that of DAVIS~ which is based upon the selective chelation of calcium by EDTA at PH 5_z5 and the removal of strontium by absorption on a cation eschange resin. Although this method was developed for the estraction of strontium from cnlcium it was found capable of reducing the barium content of analytical grade calcium carbonate from 25 to 2 p.p,m. A correction was mnclc for the barium content of the matris when synthetic standards were prepared. NETHOU

The bone ash is grouncl with a spectrographic buffer composed of equal parts of anhyclrous copper sulphatc and graphite powder. ‘I‘lic gxaphitc contains 0.5% of lanthanum, in the form of osiclc, as internal standard. l’cllets prepared from this mixture are burned to completion in graphite cups at IOA (l.c., the arc being surrounded by a mantle of osygen/argon. Spectra arc evuluatcd by non-recording micropllotometry, calibration being carried out by means of an iron intensity pattern.

Large Glass Spectrograph, de Gramont Arc and Spark stand, condensing lenses (see Fig. 3) ; sighting lamp and mask/screen, quartz filter (320/Otransmission), N.C.C. sllf

I’ 0. I I 0 u I ! I I l-l stcpktor F 1084

OVePSll Filter

Mask

4mmGap

F958

ARC

I

Distoncc Slit. In cm from

2

10

22

36

. 50

(gmdc I) graphite electrodes I/G in. diameter, 7 mm clinmctcr copper clcctrodcs, 5 mm cliamcter iron electrodes, non-recording microphotometcr, rotating 7 step sector Stallwood jet, flow gauges (2, o-5 l/min). Sfiectrogvajhic

conditiom

Spectrograph, large glass; cstcrnal optics, see Fig. 3; plate mask, 3 mm; slit length, x.8 mm ; slit width, 0.015 mm ; Wi~VClCll~tll range. 4000-5500 A ; optical filter, overall neutral quartz filter 32% transmission ; photographic plate, Ilford ordinary; top elec troclc (- vc), x/4 in. dinmctcr N.C.C. graphite 3o” cone; bottom clcctrocle (+vc), x/4 in. diameter N.C.C. graphite specially machined (see Fig. 4) ; analytical gap, 4 mm ; sample loacl, pellets prepared in special block (see Fig. 5) ; current, IOA d.c. ; esposurc, complete burn (indicated by the arc becoming unstable and noisy); Stallwood jet, see Fig. x ; gas flow rates, osygcn I l/min, argon 3 l/min. Plate

calibration

s~ectrunt

After completing esposure of samples or standards, photograph on the same plate an intensity pattern by means of a y-step rotating sector (step ratio I : 2). Spectrograph, large glass; position of sector, at slit; top electrode, (+ve), 7 mm copper, 80” I’

Aural.Glrina. klcfa,26 (1962) 349-354

M. S. .W. WEBB,

352

M. L. WORDINGHAM

cone ; bottom electrotlc (- ve), 5 mm iron, flat ; slit length, 12 mm; plate mask, 12 mm; analytical gap, 3 mm; current, 3A d.c. ; pre-arc, 30 set; exposure, 40 sec. Photographic

processing

Develop in I.D. z for 4 min at zoo. Rinse with water and fix in acid hypo until the plate is completely clear. Wash for 30 min in an efficient washing tank and allow to dry. Drying may be accelerated by soaking the plate, after washing, in 80% mcthylated spirits for 30 sec.

n I

50mm

I

Lk

Standards Prcparc by dry grinding in an agate mortar a mixture of barium sulphate and barium-free calcium phosphate such that the final matris contains IOO p.p.m, barium. I3y successive dilution prepare standards containing 30, 20, IO, 5, z and I p.p.m, Sfiectvographic

buffer~internal

standard

Prepare by dry grinding in an agate mortar for IO min equal quantities of anhydrous copper sulphate and internal standard mixture. The anhydrous copper sulphatc should be freshly prepared by heating to constant weight at 250°, and stored in an air-tight container. Prepare the internal standard misture by clry-grinding ignited lanthanum oside with N.C.C. graphite powder such that the final matrix contains 0.5% lanthanum. Pre#aration

of sam+le

Grind in an agate mortar,

for 5 min,

IOO

mg of sample with

200

mg of spectro-

SPECTROGRAPHICDETERNINATION

Ba IN BOKEASH

OF

353

graphic buffer/internal standard. Prepare 5 pellets from this mixture by loosely filling the hole in the die-plate of the pelleting block (see Fig. 5), placing the punch in position and giving a tap with a light hammer. If the barium content is grcatcr than 25 p.p.m. in the matris, dilute the sample with barium-free calcium phosphate so that the barium content falls within the range 2-25 p.p.m. Spectrogra#Gzic

pocedure

Place a pellet into the anode cup and using a clean insulated graphite rod, strike an arc between the electrodes. Allow to burn to completion at IO A, the clcctrodes being adjusted to maintain a constant gap. Rcpcat three times to give quadruplicatc esposures.

Esaminc the spectra qualitatively by comparison parator to identify and mark the following lines: I3n II *[SS).O A,

with a standard plate in a com-

1-a II 4558.5 A.

PC 4547.9 A

Measure the Scidel densities of the barium and lanthanum steps of the plate calibration spectrum.

L

I

5

10

lines and iron line in the

I

I

13

20

23

~.am.Ba InASH Fig. G. Stantlartlcurve.

Icy means of the curve obtained by plotting Seidel density of the sectored iron line against log relative intensity (from the known step ratios), convert the Seidcl densitics !of barium and lanthanum to log relative intensities, and obtain the relative intcnsity ratio barium/lanthanum. Read the concentration of barium from the standard curve (see Fig. 6). Standardisation Synthetic c

($wejaratiort standards

of standard

are prepared

curve)

by dry-grinding

barium sulphate with Anal.

Chim.

Ada,

20

calcium

(rgGz) 349-354

h1. S. W.

354

WEBB,

M. t.

WORDINGHAM

phosphate, and these are treated as above. The standard curve is obtained by plotting the relative intensity ratio barium/lanthanum against concentration (see Fig. 6). ACCURACY

AND

PRECISION

precision of the method was clctcrmined by the replication on five plates of a of variation was sample of bone ash containing 8 p.p.m. of barium. The coefficient 8% for single exposures. As a cheek on accuracy, samples were examined by the neutron activation technique, and no significant bias was evident, as will he seen from Table I.

The

‘1’/\131,15I

---__-

--_ />.p,m. Iifdritcm i*r cash

--.

__...-I:‘urissirm

.ssutpl~ No.

NlWf VIIU earl ivd iou

S/JLTl~fW~~py

WA_ I

4.6 5.0 5.8 17.3 7.‘) 4.7

5.0 5.” 5.0 ICJ.0 g.0 5.0

9 4 5 0 --_-________-

--

hCI~NOWI,I~DGlihll’N~

The assistance given by Dr. G. HARRISON of the Medical Rcscnrch Council, liacliobiological Unit, Harwcll, who proviclccl samples, the barium content of which had been detcrmincd by neutron activation analysis, is gratefully acknowlcdgcd. SUXl~lI\I
buffer mncl ~iLll~~lilJllllll oxide ilS internal stu1tkirtl. burnt in a (l.c. arc surrountlctl l)y il lllirl’ltll2 of non-rccortling iiliCrc.Jl’litrt~liIctry. ‘I’lic cffcctivc th2 ;Udl, ;Ultl lllo ccwfficicnt: of Vi~riirtioll is HI%,

ctrlqwr sulpliiitc ‘1’11~mixtur-c is pwSSc(l illto ‘I’hc spectra osq’fqm and trrfpl. c~lJlCX!lltlX~iOll

single

for

I< I’iSlJAl DOSirgC!

s~~cctroRri~~‘llicluc

:llbl;rn& clc

I’oxyclc

h

till

gwpliitc

tic kktlianc

tlU ct

cl11 SUlfirtL!

coininc

tot;al

I~i~~y~ll~~ tic

cUivrc

dt:ilo~l intcrno.

ClilllS

as slxxtrographic 30-111g

pcllctu

ax-L: cvaluntccl is r-25 p.pm. c~f hriurn at tllc 8-p.p.111. Ic\*cl

r;wf+

cspmircs

iud

lay

in

I:: Ii1

ccnclrc

anlrsdro,

tl’or;:

C~~IIIIC:

Ixs spcctrcs

sent

I,‘bcllilntillolI

tirlnpon

clbtxrniinJs

h

iUlnly!Wr

sl’cctrcJRr;rI’lricluc,

C2St ct

par Illicrol’liotoiilbtric.

%lIS,\~l31I~KPASSIJNC; %llr ~~CStinllllUll~ clCS ~~;rrillllls ill thnnoxytl XII ‘L‘aIdctt.cn vcrprcsst;

~~ll~Jl2h~llilS~hl.!

uncl

clic Spcktrcn

wird

ciicsc

inif:

Grirphit,

Iiiikr~~l’hotonlctrisch

Iirlpfcrslll

ausgcwcrtct.

Tat

untl

IAll-

h