Henryk Zbigniew WREMBEL THE DETERMINATION OF MERCURY
Transkrypt
Henryk Zbigniew WREMBEL THE DETERMINATION OF MERCURY
Henryk Zbigniew WREMBEL U niversity of Education in Słupsk THE DETERMINATION OF MERCURY AT ULTRAMICROTRACE LEVELS IN WATER USING THE LOW-PRESSURE RING DISCHARGE C o n t e n t s : 1. Introduction, 2. Method, 3. M easuring point, 4. A pplication lim its of the method, 5. E ffect of w ater com ponents on the results, 6. D eterm ination of m ercury in the B altic Sea w aters, 7. Conclusions; Streszczenie; References. 1. INTRODUCTION The ability of m ercury to bind w ith plant and anim al proteins favours accum ulation of th e elem ent in living organisms. The bioconcentration of m ercury occurs p articu larly efficiently in the m arine food chains. In fish tissue the concentration of m ercury by accum ulation can be by 3—5 orders of m agnitude higher than th a t in sea w ater. The m ercury content in m arine p red ato ry fishes can be estim ated from the following relationship [8] provided th a t th e m ercury level in sea w ater is known: CF = 0.013 Cw - 0.179, w here CF is the concentration of m ercury in fish, in mg kg"1, and Cw is the concentration of m ercury in w ater, in mg d m '3. The m ercury content in m arine p red ato ry fishes can attain high values (Fig. 1), even in w aters w ith relatively low concentration of m er cury. The bioconcentration of m ercury in fish is m arkedly m ore effi cient at higher levels of the elem ent in w ater. M any papers have been confined to th e problem of assaying low levels of m ercury in w ater. Those which constituted th e tu rn in g point w ere those by Stock and Zim erm ann [16], who developed the m icrom etric m ethod for assaying m ercury, and by H atch and O tt [11] who applied flam eless atomic ab sorption spectrom etry for th e determ ination of m ercury in w ater. This method, together w ith various m odifications of th e procedure, has found w idespread application to assay m ercury in w ater at u ltram icrotrace levels (10"8 < C ;. < lO-5 g dm"3) [2, 5, 19]. However, recently intensive studies have been carried out to assay m ercury in w ater at subm icroOceanology No 13(1980) Fig. 1. Mercury contemt in m arinę fishes vs. cancemtration of mercuiry in sea w ater (after [8]). 10 ¿0 60 -3 Hçj co n ce ntra tio n in w a te r C ^ .n g .d rn 20 30 50 70 C» Rys. 1. Zawartość rtęci w rybach m orskich w za leżności od stężenia rtęci w morzu (według [8]) trace levels (10*11 C ^-^IO '9 g dm"3) by other m ethods [1, 12, 13], At these levels the m ethod of flam eless atomic absorption spectrom etry has found only lim ited application [2 ,14]. This paper shows the possibility of using th e atomic emission spec trom etry of a low -pressure ring discharge for assaying m ercury in w a ter. The ring discharge utilized in this w ork occurred over th e pressure range 13 — 1 333 P a (0.1 — 10 Tr), hence u nder the term ,,ring dischar ge”, this particu lar kind of discharge is m eant. The restriction of this term appears to be perm issible, since the high-pressure kind of ring discharge has commonly been term ed „inductively-coupled plasm a” [7, 9, 18], Owing to im provem ent of the m ethod and exchange of some ac cessories, the sensitivity of the m ethod increased and th e detection lim it of m ercury in w ater changed as com pared w ith earlier resu lts [20, 21], 2. TIIE METHOD M ercury was assayed in v/ater by m easuring th e intensity of th e Hgl lines of m ercury selected from the emission spectrum of th e low -pressu re plasmoid of th e ring discharge. Fig. 2. Block diagram of the mercury concentration m easuring point. HF — h igh-frequency generator; I — inductance coli (exciter); S — spectral lamp: M — monochromator; Ph — photomiultiiplier; R — recorder Rys. 2. Schem at blokow y stanow iska pom iarowego służącego do oznaczania rtęci. HF — generator drgań elektrom agnetycznych, I — cew ka indukcyjna (wzbud nik), S — lamtpa spektralna, M — m onochromator, Ph — fotopow ielacz, R — re jestrator To ru n the determ ination, a 1-cm3 sample of the w ater to be analy sed was vacuum -dried at 243 K (-30°C) in the discharge cell of th e spec tra l lamp. A fter drying the sample, the lamp was evacuated to a pressu re of 13 P a (0.1 Tr) followed by switching on the generator of high-frequency electrom agnetic field to cause ignition of the plasm oid of the ring discharge in th e discharge cell of the spectral lamp. The radiation of the plasm oid was directed (Fig. 2) into the inlet slit of the m onochro m ator. The m ercury spectral line isolated from the radiation of the pla smoid was directed to a photom ultiplier and the generated photocurrent to a recorder cr electrom eter through an am plifier. The m ercury con te n t of the sample was read out from the recorder or electrom eter. The low -pressure ring discharge, som etim es referred to as the H -type discharge, is generated w hen the discharge cell containing a ra refied gas is placed in the alternating m agnetic field of an inductance coil (providing an exciter) coupled w ith a high-frequency generator. The spectrum of the plasm oid of th e low -pressure ring discharge car ried out in a spectral lamp filled w ith air at a low ered pressure of the order of 13 — 1 333 P a (0.1 — 10 Tr) has a rich stru ctu re over the visible and near-ultrav io let regions [20, 21], The stru ctu re is made up of the lines of atoms of the elem ents being the com ponents of air and its im purities. In the ring discharge m ercury is excited v ery strongly and in the spectrum of th e plasm oid all ultim ate lines of the elem ent appear. W hen the air in the lam p is replaced by helium , the m ercury lines are strongly amplified. The content of m ercury was determ ined photom etrically based on the intensity of photocurrent generated in the photom ultiplier. The m e asurem ents w ere ru n for the atomic lines of m ercury, H gl of 253.62 and 435.83 nm. As the la tte r was b etter fitted to the region of the m axim um sensitivity of the photom ultiplier, it proved to be th e most suitable for the determ inations. This was particu larly adventageous w hen using a spectral lamp filled w ith helium as a buffer gas. A part from th e m ea surem ents of the intensities of the 253.62 and 435.83-nm lines accompli shed by m eans of the photom ultiplier, visual control observations (by means of a spectroscope) w ere made of the 546.07-nm Hgl line, which was very strongly excited in the spectral lamp employed. 3. MEASURING POINT The m easuring point (Fig. 3) was assembled by using generally accessi ble analytical and m eansuring instrum entation. A few specially — con structed devices provided the exceptions. The appliances consisted of the following: (i) S pectral lamp. A specially-constructed electrodeless spectral lamp was used, the schem atic of which is shown in Fig. 4. It consists of a dis charge cell and a connector. The discharge cell was in th e form of a quartz (Vitreosil) cylindrical cuvette (150 mm lang X 11.5 mm i.d.) Fig. 3. Schem atic diagram o f th e point used for m easuring th e concentration of m ercury in ¡sea w ater. HF — generator of electrom agnetic vibrations; I — exci ter; O — oil ejector; P — vacuum pump; S — spectral lam p; Tr •— vacuum meter; Sp — spectroscope; L — quartz lens; M — monochromator; Ph — photom ultiplier; R — tape recorder; E — electrom eter; VC — digital voltm eter Rys. 3. Schem at stanow iska pom iarowego użytego do oznaczania rtęci w w o dzie m orskiej. HF •— generator drgań elektrom agnetycznych, I — wzbudnik, O — odrzutnik oleju, P — pompa próżniowa, S — lam pa spektralna, Tr — próżniom ierz, Sp — spektroskop, L — soczewka kw arcow a, M — m onochromator, Ph — fotopow ielacz, R — rejestrator, E — elektromeitr, VC — w oltom ierz cyfrow y Fig. 4. Schem atic diagram of the spectral lamp; 1 — valve cutting off the vacuum system ; 2 — con n ection w ith the vacuum system ; 3 — connector; 4 — vacuum joint; 5 — plasm oid of the ring discharge; 6 — inductance coil (exciter); 7 — filter; 8 — valve for filling the pump Rys. 4. Schem at użytej lam py spektralnej; 1 — zawór odcinający układ próżniow y, 2 — połączenie z układem próżniow ym , 3 — łącznik, 4 — szlif próż niow y, 5 — plazm oid w yładow ania pierścieniow ego, 6 — cew ka induktora (wzbudnik), 7 — filtr, S — zawór służący do napełniania pompy The connector, equipped w ith a vacuum valve to cut off the lamp from the vacuum system and another valve to aerate the lamp, was made from P y rex glass. The discharge cell was attached to the connector by m eans of a vacuum jonit; (ii) Spectrom etric system. This consisted of a prism atic m onochro m ator (SPM-2, C arl Zeiss in Jena), the spectroscope (UM-2, Unimash) for visual observation of the 546.07-nm line and the photom ultiplier (M 12 FQS 35, W erk fü r Fernsehelektronik); (iii) The vacuum system. This consisted of a rotation vacuum pum p (BL-82, UNITRA), oil ejector (of own construction), and a constant-resistance vacuum m eter (PSO-69, ZOPAP); (iv) The recording system, which included a tape recorder (K-201, Carl Zeiss, Jena), an electrom eter (219A, UNITRA)' and a digital m u lti m eter (VC-10T, UNITRA); (v) Feeding system of the spectral lamp. A specially-constructed high-frequency generator (f = 62 MHz, P out = 8 4 W) was used. The generator was fed w ith an a.-c. 220-V voltage stabilized to w ithin 0.01%. The m easurem ents w ere perform ed at a slitw idth of 0.025 mm and slit height of 20 mm. 4. APPLICATION LIMITS OF THE METHOD Calibration graphs for th e m easuring system w ere constructed separa tely for distilled w ater, artificial Baltic Sea w ater and artificial sea w ater. A rtificial Baltic Sea w ater (Bz) was prepared by em ploying the procedures reported in [10, 17], w hereas artificial sea w ater (S2) was prepared according to the Polish S tandard Artificial sea water [23], In both kinds of sea w ater the suppression of some m ercury lines was ob served. The suppression was relatively strong for the 253.62-nm line. This was m ostly due to the interaction of radiation v/ith some atoms of the salts dissolved in the w ater, m ainly w ith chlorine (Cl) and p a r tly w ith chlorine m olecules (Cl2) The filling of the spectral lamp w ith helium resulted in a very strong intensification of th e atomic lines of m ercury. In th e presence of helium th e intensity of the 435.82-nra line increased considerably. In Fig. 5 a calibration graph is show for samples of distilled w ater polluted w ith m ercury, constructed for the 253.62-nm line. The calibration graph of the m easuring system obeyed th e followig relationship for th e m easuring point and the m ethod employed: JJ = a C5 l e - kCJ> where J } is the photocurrent, Cj =10~J g dm"3 is the concentration of m ercury in w ater, and a, b and k are experim ental constants. The 7 — O ceanologia i- -l.'trif c s i .Aj . ..r U . ... s. b . n P h o tom u ltiplier Fig. 5. Calibration graph of the m easuring system foir distilled waiter Rys. 5 K rzyw a cechow ania układu pom iarow ego dia w ody destylow anej response J j , n A Fig. G. Calibration graphs of the .me asuring .system for th e ranges of u ltramicro and subm icrotrace concentrations of mercury: D —■ in distilled w ater, B z — in artificial Baltic Sea waiter, S z — in artificial sea water. Ryis. 6. K rzyw e cechow ania układu pom iarow ego w zakresie uiltramikro i subimikrośladowych .stężeń rtęci: D — w w o dzie destylow anej, B z — w zastępczej w odzie bałtyckiej, Sz — w w odzie mor skiej coefficient, k, describing self-absorption of the radiation in the plasmoid, am ounted to 0.037. For the u ltra — and subm icrotrace levels of m ercury in w ater (10_8^ C y. =d0"e and 10'n,^Cy. <10~9 g d m '3, res pectively), th e attenuation of the signal due to self-absorption is n e gligible (Fig. 5) and consequently the expression for the calibration graph assumes the form: Jj = a C 1In the distilled w ater th e values of a=2350 and b = 0.23 w ere ob tained over this concentration range of m ercury for th e 253.62-nm line. Due to attenuation of the 253.62-nm line by some com ponents of the sea w ater, the calibration graphs for particu lar kinds of the w ater differred. The sensitivity of the m ethod was also low er for th e two kinds of sea w ater than for distilled w ater. In Fig. 6 calibration graphs of the system are shown over the u l tram icro trace concentration range of m ercury for the th ree kinds of w ater. The detection limit, CL, was determ ined from the equation reported in [6]: JL = Jn + s j / | ' [tL + t N) w here J L and J N are th e signals corresponding to th e detection lim it and noises, respectively; s = m ax. (S i , s#), w h ere sL an d are s ta n d a rd deviations of the signal and noise, respectively: t L and t N are respec tive critical values of th e t variable of the S tudent distribution. The determ ination limit, CD, of m ercury in w ater was determ ined as a concentration at which th e erro r of the m ethod am ounted to 25 per cent. It w as calculated from th e expression reported in [13]: J d = Jn + w here J D and J N are signals corresponding to the detection lim it and noise, respectively. Table 1 lists num erical values obtained for the detection and de" term ination lim its for various kinds of w ater. Table 1. D etection lim it (C7 ) and determ ination lim it (CD ) of m ercury in various kinds of w ater (D — distilled water; B . — artificial B altic Sea water; S z — arti ficial sea water) Tab. 1. Granica w ykryw alności (C L ,) i oznaczalności (C D ) rtęci w różnych odmia nach w ody (D — woda destylow ana, B j — zastępcza woda bałtycka, S* — zastęp cza woda morska) Kind of w ater Odmiana wody CL, mg dm-! C d , ng dm -3 D 0.001 0.3 Bz 0.01 2.0 0.2 10.0 A series of control determ inations confirm ed the suitability of the developed m ethod and technique for assaying ultram icro trace am ounts of m ercury in sea w ater and subm icrotrace am ounts of the elem ent in distilled w ater. The results of these determ inations are listed in Table 2. The m ethod and m easuring technique developed in this w ork en abled to shift the determ ination lim it of m ercury in w ater at least by one order of m agnitude as com pared w ith the flam eless atomic ab sorption spectrom etry. By using this m ethod and technique it is possibTable 2. R esults of control m easurem ents in various kinds of w ater (D — distilled w ater; B . — artificial B altic Sea water; Sz — artificial sea water; SR — relative standard deviation; n = 20) Tab. 2. W yniki pom iarów kontrolnych w różnych odmianach w ody (D — w oda destylow ana, Bz — zastępcza w oda bałtycka, — zastępcza woda morska, SR — w zględne odchylenie standardowe, n = 20) Kind of w ater Odmiana w ody D Bz Sz Hg added Dodano Hg ng dm-5 0.00 100.00 0.0 100.0 0.0 100.0 H g f o u nd Wylkryto Hg n g dm-3 0.1 97.6 0.4 92.7 0.5 89.4 S r 0.09 0.12 0.17 le to assay m ercury in w ater over th e range betw een ultram icro and submicro trace levels. 5. THE EFFECT OF WATER COMPONENTS ON RESULTS To determ ine the effect of sea w ater com ponents on th e signal in ten sity, a series of m easurem ents w ere ru n for the 253.62-nm line w ith populations of samples of nom inally equal m ercury concentrations, p re pared by using solutions containing some com ponents of the artificial sea w ater only. Solution Ri contained only sodium compounds which make up approx. 79.4 p er cent by w eight of all com ponents of artificial sea w ater. Solution R2 contained th e rem aining com ponents of the w ater, w ith the exception of sodium compounds, w hich constitute about 20.6 per cent of all components. Solution R 3 com prised heavy m etal n itrates w hich occur in artificial sea w ater at a level of 0.0005 p er cent. As regards samples prepared using distilled w ater, suppression of the signals was observed-in these solutions (Fig. 7). In an extrem e case, the attenuation of the signal in artificial sea w ater of full composition am ounted to approx. 65 per cent. In the rem aining solutions, the stron gest attenuation was noted in solutions containing chlorine and sodium compounds. O ther sea w ater com ponents did not affect th e signal intesity appreciably. Fig. 7. S ignal in ten sities fo rssolutions of various contents of sea w ater com ponents (norma lized to signal 'in distilled w a ter) at various concentrations of m ercury in water. Rys. 7. W ielkość sygnału dla roztw orów o różnych zaw artoś ciach składników w ody morskiej (unorm owane do sygnału otrzy m yw anego dla w ody destylow a nej) przy różnych ¡stężeniach rtęci w w odzie. 6. DETERMINATION OF MERCURY IN BALTIC SEA WATER The concentration of m ercury in Baltic Sea w ater was m easured by this apparatus and m ethod in 1976— 1978. W ater samples w ere taken in the coastal zones of Kołobrzeg and U stka and in th e region of th e Słupsk Sandbank (Ławica Słupska). A few samples taken in th e G ulf T able 3. The content of m ercury in som e B altic Sea regions (C = = concentration o f m ercury in water); c = m ean concentration of m ercury in water; S R = relative standard deviation; n = number of m easurem ents) Tab. 3. Zawartość rtęci w niektórych obszarach Bałtyku (C — stę żenie rtęci w w odzie, c — średnie stężenie rtęci w wodzie, S r — w zględne odchylenie standartowe, n — liczba pomiarów) R egion Rejon C, ng dim-5 c, ng dm-3 Sr n C oastal walterls (Ustka) Wody przybrzeżne 20— 180 40 0.75 30 C oastal waiters (Kołobrzeg) W ody przybrzeżne 30— 190 45 0.82 30 10—>150 34 0.45 12 120—360 105 0.67 7 Słupsk Sandibank Ław ica Słupska G ulf o f Gdańsk (Sopoit) Zatoka Gdańska of G dańsk w ere also analysed. They w ere stored in polyethylene con tainers and th e analyses w ere ru n 24 hrs. after sampling. The sam ples w ere acidified w ith n itric acid to pH<[2 and 0.1 g of potassium io dide was added per dm 3. N itric acid was added to minimize losses of m ercury due to sorption on th e w alls of the container, w hereas po tassium iodide decreased m igration of m ercury from th e solution to the atm osphere. The results are shown in Table 3. There w ere considerable fluctuations in the concentration of m er cury in the Baltic Sea w aters. Sim ilar resu lts from other regions of the Baltic Sea w ere reported in [3], E xtrem al concentrations of m er cury in these regions differred by tw o orders of m agnitude. A too short observation period excluded th e possibility of revealing the cyclic nature of variations of the m ercury concentration. 7. CONCLUSIONS The m easurem ents carried out in this paper confirm ed the suitability of the m ethod for assaying u ltram icrotrace levels of m ercury in sea w ater. The m ethod is p articu larly useful in investigations of subtle processes of exchange of m ercury betw een the solution and walls of vessels used in analytical procedures, as w ell as betw een th e atm os phere and w ater surface. These processes provide the main sources of vitiation of the results in the analysis of ultram icrotrace am ounts of m ercury in w ater [22], The results of the m easurem ents of the concentration of m ercury in the Baltic Sea w ater showed th a t in the regions studied the m ercury content ranged from 30 to 60 ng d m '3, occassiona'ily attaining levels of 10 and 100 ng dm"3 (in extrem al cases even more than 350 ng dm"3). The concentration of m ercury in th e Baltic Sea exhibits seasonal va riations. It w as also found th a t the sensitivity of th e m ethod was largely affected by proper fitting of the analytical line of th e elem ent to the m axim um sensitivity range of th e photom ultiplier, as w ell as the use of proper bufferring gas for filling the lamp. The use of helium as the bufferring gas increases th e sensitivity of the m ethod in the case of m ercury. Wyższa Szkoła Pedagogiczna w Słupsku OZNACZANIE MIKROŚLADÓW RTĘCI W WODZIE Z UŻYCIEM NISKOCIŚNIENIOWEGO W YŁADOWANIA PIERŚCIENIOWEGO Streszczenie Zdolność rtęci do w iązania się z białkiem sprzyja akum ulow aniu się jej w or ganizm ach żyw ych i biokoncentracji rtęci w łańcuchach pokarmowych. Biokoncentracja rtęci przebiega szczególnie efek tyw n ie w wodnych łańcuchach pokarmo w ych. W tkankach ryb m oże nastąpić zatężenie rtęci, spowodowane akum ulacją, naw et o 3—5 rzędów w ielkości w porównaniu z zawartością rtęci w wodzie. P o woduje to konieczność oznaczania naw et bardzo niskich zaw artości rtęci w w o dzie. W pracy przedstawiono m ożliw ość użycia em isyjnej spektrom etrii n iskociśn ie niow ego w yładow ania pierścieniow ego do oznaczania ultram ikrośladów rtęci w wodzie. Osiągnięta granica oznaczalności w ynosząca dla w ody destylow anej ok. 0,1 ng. dm-3 Hg/H20 , k w alifiku je się m etodę zw łaszcza do badania subtelnych pro cesów w ym iany rtęci m iędzy roztworem a otoczeniem , np. m iędzy roztworem a ściankam i naczyń, stosow anych w technice analitycznej lub do badania pro cesów w ym iany rtęci m iędzy atm osferą i powierzchnią wody. Zjawiska te są w zakresie ultram ikrośladów głów nym i źródłami błędów oznaczeń rtęci w roz tworach wodnych. Badania w ykonane w wodach bałtyckich w ykazały, że zawartość rtęci, w ob jętych badaniami obszarach morza, mieścji s ię naogół w granicach od 30 do 60 ng. dm-J, osiągając czasam i w artości rzędu 10 ng. dm-*, a niekiedy powyżej 100 ng. dm-5, w skrajnych przypadkach naw et powyżej 350 ng. dm-3. Stwierdzono również, że dla czułości m etody duże znaczenie ma dopasowanie linii analitycznej oznaczanego pierw iastka do zakresu m aksym alnej czułości w id m owej fotopow ielacza użytego do detekcji prom ieniowania, jak rów nież u życie od powiedniego gazu buforowego służącego do napełnienia lam py spektralnej. U życie helu jako buforu powoduje przy rtęci zw iększenie czułości metody. REFERENCES 1. B o u d i n G., A n alytical Methods A pp lied to W ärter Pollution , J. Radioanal. Cheim., 1977, 37, p. 123 — 139. 2. B r e i id e n b a c h A. W., C r o w e R. E., Methods for Chemical Analysis of W ater and Wastes, U. S. E nvironm ental Protection A gency, W ashington 1974. 3. B r i i g ' i m a n n L., Some R em arks on the Content of M ercury in the Baltic, Proceedings of the X I Conference of the B altic Oceanographer«, Rostock 1978, 1, p. 249 — 263. 4. B r z e z i ń s k a A., Trace Metals in the W ater of the Gdańsk Basin, Procee- 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. dings of the X I Conference of the B altic Oceanographers, Roistock 1978, p. 264 — 269. F ö n s t n e r U., M ü . l l e r G., S ch w erm etalle in Flüssen und Seen, Springer, B erlin 1974. G a b r i e l R., A General M ethod for Calculating the Detection L im it in Che mical Analysis, Anal. Chem., 1970, 42, p. 1439 — 1440. G a j e w s k i W., (Ed.), Zagadnienia p o d sta w o w e spektraln ej analizy atomo w ej, WNT, W arszawa, 1972. G a r d n e r D., M ercury in Fish and W aters of the Irish Sea and other U nited K ingdom F ishing Grounds, Nature, 1978, 272, No. 5648, p. 49-51. G r e e n f i e l d S., M c G e a c h i n H. Ms;D., S m i t h P. B., Plasma Emission Sources in A nalytical Spectroscopy, Talanta, 1976, 33, p. 1 — 14. . H o r n e R. A., Marine Ch em istry, W iley, N ew York 1969. H a t c h W. R., O t t W. L., D eterm ination of S ub-m icrogram Quantities of Mercury b y Atom ic Absorption S p ectro m e try, Anal. Chern., 1968, 40, p. 2085 — 2087. L i e s e r K. H., N e i t z e r t V., Determination of Trace Elements in Water b y N on-destructive Neutron Activation, J. Radioanal. Chem., 1976, 31, p. 397 — 405. M o e n k e H., A tom spektroskopische Spurenanalyse, A kadem ische Veiri'äSsgesellsch aft, Leipzig 1974. M u z y k o v G. G., C h u v stv ite ln o st’ opredeleniya rtu ti atomno-absorptsionn y m m e to dom izholodnogo para, Zhurn. Priklad. Spektoroskopii, 1975, 23 (5), p. 763 — 767. R e i c h e r t J. K., Spurenan alytik in G ew ässern m it Hilfe der Atomem ission, A tom absorption und Atmofluoreszenz, Vom W asser, 1975, 40, p. 135 — 149. S t o c k A., C u c u e l F., Die Verbreitung des Quecksilbers, N aturw iss., 1934, 22, p. 390 — 393. T r z . o s i ń s k a A., P o d s ta w o w y sk ład jo n o w y w o d y bałtyckiej, Studia i M ate riały Oceanologiczne, 1977, 17, p. 165 — 180. W i n e f o r d n e r J. D., Trace Analysis — S pectroscopic Methods for Elemen ts, J. W iley, N ew York 1976. W i n k l e r H. A., Vorkom m en von Quecksilber in G ew ässern und seine ana lytische Erfassung, Chemie Ingenieur Technik, 1975, 47 (16), p. 659 — 666. W r e m b e l H. Z., Zastosowanie e m isy jn e j analizy w i d m o w e j do w y k r y w a n i a śladow ych ilości rtęci w wodzie, S tudia i M ateriały Oceanologiczne, 1977, 17 p. 415 — 438. W r e m b e l H. Z., Z astosowanie o p tyczn ej analizy em isy jn e j do w y k r y w a n ia śladow ych ilości rtęci w wodzie, Studia i M ateriały Oceanologiczne, 1977, 19, p. 128 — 135. W r e m b e l H. Z., The Use of Ring Discharge S p e c tro m e try for Some In ve stigations of the Behaviour of Mercury in W ater, P roceedings of the XT Con ference of the Oceanographers, Rostock, 1978, 2, p. 679 — 691. Z astępcza woda morska, Polska Norma PB-66/C -06502, 1966.