ISSN:0003-2654    

DOI:10.1039/AN96186FX029

出版商:RSC    

年代:1961

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96186BX031

出版商:RSC    

年代:1961

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96186FP141

出版商:RSC    

年代:1961

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96186BP151

出版商:RSC    

年代:1961

数据来源: RSC

摘要:

JULY, 1961 THE ANALYST Vol. 86, No. 1024 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY DEATHS WE record with regret the deaths of Edgar Percy Campbell Albert Edward Fletcher Harold Wright Hodgson Reginald Arthur McNicol. 429 The Toxicology of Nitrates and Nitrites with Particular Reference to the Potability of Water Supplies A Review* BY E. H. W. J. BURDEN? (Firthorfie, South Ascot, B e h . ) INTRODUCTION THERE have been many reports in recent years that high concentrations of nitrates in drinking supplies have caused poisoning and death among infants and also among animals. A review has been made of the evidence available in order to suggest suitable limits for this constituent in drinking water supplies. THE TOXICOLOGY OF NITRATES AND NITRITES MODE OF ACTION- After ingestion, nitrates are converted to nitrites by bacterial reduction.According to Cornblath and Hartrnann,l this reduction occurs in the lower intestine of the adult. However, in very young babies, which have a physiological gastric achlorhydria, reduction occurs in the stomach and duodenum from which the nitrites are more readily absorbed into the blood stream. After absorption, nitrites convert oxyhaemoglobin into methaemoglobin and thus interfere with the oxygen transport in the blood. This effect may be cumulative, since the methaemoglobin is reduced only at a slow rate. * Reprints of this paper will be available shortly. t Formerly The Government Analyst, Ministry of Health, Khartoum, Sudan. For details, please see p. 492.430 BURDEN : TOXICOLOGY OF NITRATES AND NITRITES WITH vol.86 They occur in the order: (a) pain in the epigastrium; (b) purging, often with blood; (c) convulsions of the face muscles; (d) weak irregular pulse; (e) difficult respiration; (f) cold extremities followed by collapse. SYMPTOMS OF POISONING- Luff2 describes the symptoms of acute potassium nitrate poisoning. Cyanosis is usually observed and t:he blood becomes chocolate coloured. ANTIDOTE- to haemoglobin, and the patient makes a dra.matic recovery.3 Intravenous injection of methylene blue results in a rapid reduction of methaemoglobin LETHAL DOSE: HUMAN- Nitrates-Smith and Simpson4 consider the normal fatal dose of potassium nitrate to be 15 to 30 g for an adult, although they quote a report by Peterson, Haines and Webstefl of the death of an adult after taking 8 g.Luff2 records the death of an adult after taking 3.5 g of potassium nitrate, although Witthaws reports a recovery after taking 125 g. C0mly7 quotes the observation of Eustennan and Keiths that certain adults are extremely allergic to nitrates. These authors record that one adult suffered from extreme cyanosis after taking 7 g of ammonium nitrate daily for 8 successive days. Wood3 also states that nitrates are a cumulative poison. On the other han.d, Ambergg reports that a 13-year-old boy (110 lb) with glomerulonephritis took 6 g of potassium nitrate daily for a year without injury. Windmuellerlo states that an adult who took 4 teaspoonsful daily of an equal mixture of sulphur and potassium nitrate for 26 days died 17 days later.Nitrites-Smith and Simpson4 consider that the normal fatal dose of sodium nitrite is about log. Manicatidell records the death of a 3-month-old child after taking 0.35g of sodium nitrite, but the recovery of a 2-month-old child after taking 0-4 g. Tankard12 reports the death of an adult who probably took about 7 g of sodium nitrite. Scholes13 records the accidental death of three adults, two of them within an hour, after eating a meal seasoned with sodium nitrite instead of sodium chloride. Several other deaths from nitrite poisoning have been r e c ~ r d e d . ~ * J ~ J ~ J ~ J ~ However, Kobert, quoted by Autenrieth,lg mentions the case of a patient who took “several doses” of 0.5 g of sodium nitrite at hourly intervals “without dangerous symptoms,” although the srymptoms included nausea, diarrhoea, cyanosis and increased diuresis.Naidu and Venkatrao20 consider 2.0 g of sodium nitrite to be a fatal dose for a 10-stone man after examining results obtained on dogs. This is equivalent to 32 mg per kg. FATAL DOSE : ANIMALS- Crawford21 reports a case of sodium nitrate poisoning in cattle; he considers that & lb of sodium nitrate is a fatal dose. For an animal weighing 500 kg this is equivalent to 450 mg per kg. Bradley, Eppson and Beath22,23 have shown that “oat hay poisoning’’ in cattle is caused by abnormally high concentrations of nitrate in the affected hay. They havealso found that 1 g of potassium nitrate per kg is a lethal dose for cattle. Death is caused by methaemoglobinaemia, 70 to 80 per cent. of the haemoglobin being converted to methaemo- globin in a few hours.Sapiro, Hoflund, Clark and QuinN have demonstrated that nitrate is reduced to nitrite by the rumen contents of cattle. has shown that, in the rumen of sheep, nitrate is reduced successively to nitrite and ammonia. With large doses of nitrate, the conversion of nitrite to ammonia is limited and the nitrite concentration in the rumen increases rapidly. After giving large doses of nitrate, the methaemoglobin in the blood is found to reach a maximum about 7 hours later. Lewis finds that the doses of nitrate and nitrite required to give 60 per cent. conversionof haemoglobin to methaemoglobin are- Sodium nitrate (added to rumen) . . . . 420 mg per kg Sodium nitrite (added to rumen) . . . . 170 mg per kg Sodium nitrite (intravenously) .. . . . . 34mgperkg Sheep are similarly affected. Oltmann and Crandal126 find that the lethal dose of sodium nitrite for rabbits is 80 to Winks, Sutherland and Salisbury2’ report that the lethal dose of sodium nitrite 90 mg per kg. for pigs is 90 mg per kg, and that doses of 50 to SO mg per kg give toxic symptoms.July, 19611 PARTICULAR REFERENCE TO POTABILITY OF WATER SUPPLIES. A REVIEW 431 The lethal doses of sodium nitrite, calculated as nitrogen, quoted above for adults and various animals can be summarised as shown below- (1) For 10-stone adult . . .. .. 33 mg per kgQ (2) For 10-stone adult . . .. .. 22 mg per kgI2 (3) Rabbits . . . . . . . . 16 to 18 mg per kgZ6 (4) Pigs . . . . .. . . .. 18 mg per kg2' (5) Sheep . . . . . ... 34 mg per kg25 (6) Dogs . . . . . . . . .. 6.4 mg per kg20 A summary of the reported lethal doses of nitrate, calculated as nitrogen, is shown below- (1) For 10-stone adult . . .. . . 34 to 68 mg per kg4 (2) For 10-stone adult . . .. .. 18 mg per kg5 (3) Cattle. . .. . . . . . . 75 mg per kg21 (4) Cattle. . . . .. . . . . 140 mg per kg23 (5) Sheep . . . . .. .. 70 mg per kg25 NITRITES AND NITRATES IN WATER SUPPLIES- In most natural waters, nitrate is the final product of the aerobic decomposition of nitrogenous organic matter. Its presence in a water supply is usually interpreted as an indication of contamination by organic matter, although this may have happened in the remote past and may not necessarily be serious. Nitrite indicates less complete oxidation and may show more recent and possibly dangerous contamination. Traces of nitrate are present in most natural waters, and it is only rarely that the concentration rises to a dangerous level.The recorded cases of poisoning caused by high concentrations in water supplies come under two headings- (a) infantile cyanosis and (b) animal poisoning. (a) INFANTILE CYANOSIS- Comly7 cites from personal experience a number of cases of infantile cyanosis where symptoms were observed after ingestion of feeds made from waters containing 64 to 140 p.p.m. of nitrate-nitrogen. Taylor28 records a large number of cases of infantile cyanosis caused by nitrates in feed ~ater.2~JOv~l s32933934 Betke and Kleiha~eI3~ also refer to a case of infantile cyanosis occurring under similar circumstances.The reasons that make artificially fed infants particularly susceptible appear to be- He quotes a number of other cases including one fatality. (i) the normal gastric achlorhydria of infants, already referred to,l and (ii) the high relative fluid intake of an infant compared with that of an adult.3 In England, a baby under 15 lb requires about 2 i fluid 02. of water per pound body weight per diem.3 In the tropics, this figure is much higher and may rise to about 10 fluid 02. per pound body weight per diem.36 In comparison, a normal adult in England requires about 1.8 pints of liquid a day.37 For a 10-stone man, this is equivalent to 6 fluid oz. per pound body weight per diem. In the tropics, the requirement would be about 1Q fluid 02. per pound body weight per diem.Waring38 suggests that waters used for preparing infant feeds should not contain more than 20 p.p.m. of nitrate-nitrogen. Ewing and Mayon-White= concur with the suggestion. The Committee of the Division of Medical Science of the United States National Research Council3g regard any well water containing more than 10 p.p.m. of nitrate-nitrogen as unfit for consumption by infants. Wood3 regards supplies containing less than 20 p.p.m. of nitrate-nitrogen as satisfactory €or infant feeding, those containing 20 to 39 p.p.m. as undesirable, those containing 40 to 79 p.p.m. as unsafe and those containing more than 80 p.p.m. as dangerous, Any apparent discrepancy between these authors' findings may be explained by possible differences in climate. (b) ANIMAL POISONING- 1000 p.p.m.of nitrate-nitrogen. Cattle-Fincheflo reports the death of 4 cattle after drinking from a supply containing Successive Sudan Government Analysts have recorded132 BURDEN : TOXICOLOGY OF NITRATES AND NITRITES WITH [Vol. 86 deaths of cattle and camels after drinking from water supplies containing a high nitrate con~entration.41,~2,~3’~~,~~,~~ In the majority of cases, the nitrate-nitrogen concentration was found to lie in the range 70 to 150 p.p.m., although three supplies containing as much as 870 p.p.m. have been re~orded.4~,*~ In a recent case, a supply found to contain 320 p.p.m. was fatal to cattle in about 14 hours after wa~tering.~~ Unfortunately, as far as the records show, the bodies of poisoned animals have not been examined by veterinary inspectors because the supplies have come from wells in remote areas, and any evidence has been lost before a post-mortem examination could be made. On the basis of available evidence, I introduced a tentative limit of 50p.p.m. of nitrate-nitrogen, which has been applied in recent years in Khartoum when making recommendations on the suitability of a particular supply.However, Ibrahim has pointed that the nitrate concentration of a water supply may fluctuate within a wide range, and he re-examines regularly any supply with a nitrate-nitrogen concentration in the range 30 to 50 p.p.m. W i l l i m ~ t t ~ ~ has also had experience of cattle poisoning in Cyprus caused by well waters containing high concentrations of nitrate. Pigs-Winks, Sutherland and Sa1isbu1-y~~ record heavy mortality among pigs fed on a soup made by cooking beef and offal in a well water containing 290 to 495 p.p.m.of nitrate- nitrogen and less than 1.5 p.p.m. of nitrite-nitrogen. The soup itself contained 80 to 285 p. p .m. of ni trit e-ni trogen . Fish-Sanborn48 states that water containing 2000 p.p.m. of sodium nitrate (330 p.p.m. of nitrate-nitrogen) is fatal to fish, whereas a concentration of 1000 p.p.m. of sodium nitrate (165 p.p.m. of nitrate-nitrogen) is without apparent effect. DISCUSSION ANI) CONCLUSIONS and animals, the following assumptions have been made- In deducing a suitable limit for nitrates in water to be consumed by infants, adults (i) that the nitrate is completely reduced to, and absorbed as, nitrite; (ii) that the lethal dose of nitrite-nitrogen for men and cattle is about 20 mg per kg; (iii) that one fifth of the lethal dose, i.e., 4- mg per kg, is the maximum daily amount of nitrite-nitrogen that can be tolerated without giving rise to toxic symptoms.When computing limits for cattle in tropical climates it should be remembered that in many places cattle are watered only once or twice a day, and that, as a consequence, large volumes are taken at a time. On the above basis the limits shown in Table I for the permissible concentrations of nitrate-nitrogen in drinking water for infants, adults and cattle in England and in the tropics have been calculated. TABLE I PERMISSIBLE LIMITS FOR NITRATE-NITROGEN I N DRINKING WATER Assumed Consumer weight, kg Infant . . 3 Infant . . 3 Adult .. 60 Adult . . 60 cow .. 500 Maximum permissible Approximate daily Region dose of nitrate-nitrogen, water intake, mg litres England 12 0.5 Tropics 12 2 England 240 1 Tropics 240 5.3 Tropics 2000 45 Permissible limit of nitrate-nitrogen in water, p.p.m. 24 6 240 45 45 These figures can only be approximate, but they give general support to the limits already suggested. They also show how very susceptible an infant artificially reared in the tropics can be to relatively small amounts of nitrates in drinking water. In areas where the water supplies have a high nitrate concentration special arrangements must be made for supplying infants with low- nitrate water. I t may also be concluded that, since the lethal doses of nitrate-nitrogen and nitrite- nitrogen for adults are not very different, most of the nitrate consumed is reduced to nitrite and absorbed into the blood stream.Thus it .would appear that the significant factor that makes infants susceptible to nitrates is the high relative fluid intake and not the gastric achlorhydria. They emphasise the variations caused by differences in climate. This is already being done in Norfolk and East S~ffolk.~July, 19611 PARTICULAR REFERENCE TO POTABILITY OF WATER SUPPLIES. A REVIEW 433 REFERENCES 1. 2. 3. 4. 3 . ti. - 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 39. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. Cornblath, M., and Hartmann, A. F., J . Pediat., 1948, 33, 421. Luff, A. P., “Text Book of Forensic Medicine and Toxicology,” Longmans, Green and Co., London, Wood, E.C., personal communication. Smith, S., and Simpson, K., “Taylor’s Principles and Practice of Medical Jurisprudence,” Eleventh Peterson, F., Haines, W. S., and Webster, R. W., “Legal Medicine and Toxicology,” Second M’itthaus, R. A., “Manual of Toxicology,” Second Edition, BalliBre, Tindall and Cox, London, Comly, H. H., J . Anter. Med. Ass., 1945, 129, 112. Eusterman, G. B., and Keith, N. M., M . Clin. North America, 1929, 12, 1489. Amberg, S., Proc. Mayo Cliwic, 1935, 739. Wjndmueller, E., J. Ayner. iVed. Ass., 1921, 77, 858; Chem. Abstr., 1922, 16, 756. Manicatide, -., Bull. SOC. Ped., Paris, 1932, 379; Chem. Abstr., 1933, 27, 5824. Tankard, A. R., Analyst, 1937, 62, 734. Scholes, A., Ibid., 1936, 61, 685.Newcomb, C., Ibid., 1936, 61, 614. -, Ibid., 1937, 62, 740. Manley, C. H., Ibid., 1945, 70, 50. Padberg, L. R., and Martin, T., J . Amer. Med. Ass., 1939,113, 1733; Chem. Abstr., 1940, 34, 7004. Becker, E., Dtsch. Med. Rdsch., 1949, 3, 900; Chem. Abstr., 1952, 46, 11442. Autenrieth, W., “Laboratory Manual for the Detection of Poisons and Powerful Drugs,” Sixth Naidu, S. R., and Venkatrao, P., Calcutta Med. J . , 1945, 42, 79; Chem. Abstr., 1947, 41, 2487. Crawford, M., Trop. Agric. (Ceylon), 1941, 96, 359; Chem. Abstr., 1942, 36, 1097. Bradley, W. B., Eppson, H. F., and Beath, 0. *4., J . Amer. Vet. Ass., 1939, 94, 541. Sapiro, M. L., Hoflund, S., Clark, R., and Quin, J. I., Onderstefioort J . Vet. Sci., 1949, 22, 351. Lewis, D., Biochem. J., 1951, 48, 175.Oltmann, T. V., and Crandall, L. A., J . Pharmacol., 1931, 41, 121; Chern. Abstr., 1931, 25, 3089. Winks, W. R., Sutherland, A. K., and Salisbury, R. M., Queensland J . Agric. Sci., 1950, 7, 1; Taylor, E. W., “The Examination of Waters and Water Supplies (Thresh, Beale and Suckling).” Walton, G., Amer. .J. Public Health, 1951, 41, 986. Ferrant, M., J. Pediat., 1946, 29, 585. Homolka, J., abstracted iia Bull. Hyg., 1949, 24, 676. Campbell, W. A. B., Brit. Med. J., 1952, ii, 371. Ewins, M. C., and Mayon-White, R. M., Iancet, 1951, 260, 931. Fawns, H. T., and Aldridge, A. G. V., Brit. Med. J., 1954, Li, 575. Betke, K., and Kleihaner, E., Dfsch. Med. Woch., 1957, 82, 1127; abstracted in Bull. Hyg., 1957, Burden, E. H. W. J., personal observation. Mottram, V. H., and Graham, G., “Hutchison’s Food and the Principles of Dietetics,” Eleventh Waring, F. H., J . Amer. Wat. Wks Ass., 1949, 41, 147, Maxcy, K. F., Wat. Pollution Abstracts, 1951, 24, 97. Fincher, M. G., Cornell Veterinarian, 1936, 26, 271; Chem. Abstr., 1936, 30, 8389. Henry, A. J., Annual Report of the Government Analyst, Khartoum, Sudan, 1945, 5. Grindley, D. N., Ibid., 1951/52, 6. -, Ibid., 1952/53, 6. Burden, E. H. W. J., Ibid., 1956/57, 9. Ibrahim, A. H., Ibid., 1957/58, 8. Willimott, S. G., personal communication. Sanborn, N. H., Canner, 1945, 101, 5 and 13; Chem. Abstv., 1946, 39, 4607. 1895, Volume I, p. 147. Edition, J. & ,4. Churchill, London, 1957, Volume 11, p. 292. Edition, Saunders, Philadelphia. 1911, p. 305. Edition, J. & A. Churchill, London, 1928, p. 321. , , , Ibid., 1940, 96, 41. - ~ - Chem. Abstr., 1951, 45, 3507. Seventh Edition, J . & A. Churchill, London, 1958, p. 548. 32, 979. Editon, Edward Arnold (Publishers) Ltd., London, 1956, p. 399. -, Ibid., 1957/58, 8. Received Janziary 16th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9618600429

出版商:RSC    

年代:1961

数据来源: RSC

摘要:

434 DEAN AND HERRINGSHAW: SOME ASPECTS [Vol. 86 Some Aspects of Pipette Usage Part I. Precision BY G. A. DEAN* AND J. F. HERRINGSHAW (Chemistry Department, Imperial College of Science and Technology, Logzdon, S. W.7) The component errors in the variability of the volume delivered are analysed and their magnitudes determined or estimated. The resultant of these is shown to be in good agreement with th.e experimentally determined precision of the volume delivered. The method used in the manipulation of pipettes is described. THE experiments described originated in an attempt to evaluate the errors in measuring volumes in a series of comparative titrations. They are not an attempt to show how precise the deliveries from a pipette can be made. This has been done by Bishopl; by observing the meniscus with a travelling microscope, by working under conditions in which the ambient temperature and pressure were remarkably uni Form, by rigidly standardising every operation involved and by taking the values for which replication was satisfactory, he obtained standard deviations of 0-0002 to 0.0003ml in the deliveries from 5- to 100-ml pipettes.Although this probably represents the maximum precision attainable with pipettes, it was not intended to indicate the precision obtainable in normal use. However, we found that the precision obtainable under ordinary conditions was remark- ably good and were prompted to investigate not: only the various contributions to the over-all precision (Part I), but also the effect of changing certain variables in technique and in conditions (Part 11).GENERAL PROCEDURE- The pipettes used were selected grade-B one-mark bulb pipettes and 2- and 50-ml pipettes (designated M2- and M50-ml hereafter) modified by collapsing the stem at the mark to form a capillary of about 1 mm inside diameter and about 1 cm long and making a new mark around the centre of this portion; the tip was drawn out to form a slender tip with thin glass walls. These modifications were designed! to reduce errors associated with the tip and in setting to the mark. All apparatus was cleaned before use with chromic - nitric acid cleaning mixture; it was then thoroughly rinsed with distilled water and finally with the liquid to be used for calibration. To avoid heating the working poirtion, the pipettes were held by the extreme top and by a 3-cm cube of plastic foam placed around the lower stem, The temperature recorded was that of the liquid in the reservoir (usually a 1-litre beaker nearly full) from which the pipette was being filled, unless (as in some special experiments) the temperature of the liquid differed markedly from that of the air, when the temperature of the liquid delivered was determined directly.The pipettes were filled by suction until the meniscus was about 2 cm above the mark and the tip was wiped with a piece of absorbent tissue. With the pipette held vertically and its tip in contact with the wetted lip of a, clean beaker, the meniscus was allowed to fall slowly until it was on the mark. The beaker was then removed, the meniscus level was checked, and the pipette was allowed to deliver freely (i.e., with the tip in air) in a vertical position, the liquid delivered being collected in a 50-ml conical flask, which was closed with a rubber bung during weighing.When flow had ceased, the pipette was allowed to drain in the vertical position for 20 seconds, after which the tip was touched against the surface of the liquid and then withdrawn. The last process occupied about 2 seconds, hence the total drainage-time is quoted as 22 seconds. The modified pipettes were emptied as described above; after 20 seconds’ drainage, the liquid in the tip was expelled with the tip in contact with the wall of the flask. For finger control of the meniscus the slightly moistened forefinger was used. For “pneumatic control,” a length of rubber tubing was fitted to the pipette and two screw-clips were placed 1 and 3 cm from the junction; with both clips fully open the pipette was filled to above the mark, the outer clip was closed and the final adjustment was made by manipu- lating the inner clip.The weights of liquid delivered were converted into volumes-at-20” C * Present address : Chemistry Department, Dalhousie University, Halifax, Xova Scotia, Canada.July, 19611 OF PIPETTE USAGE. PART I. PRECISION 435 by means of the Tables given by Strouts, Gilfillan and Wilson.2 The distilled water used was not boiled out-although boiled-out water is recommended by many authorities it is difficult to see how such water can be used without it partially reverting to its original state.Weighings were made with calibrated weights, and were recorded to 0.1 mg. PRESENTATION OF RESULTS- Repro- ducibility is expressed as a standard deviation and is followed (in parentheses) by the number of determinations involved. The volume delivered, corrected to 20" C, is designated volume D 20" C in the Tables. Volumes are calculated generally to 0.0001 ml, whether this is significant or not. Fig. 1. Meniscus a t a graduation1mark PRECISION OF SETTING TO THE MARK- If a meniscus in perfect tangential contact with the top of a graduation line is allowed to move beyond this position, a region of mutual overlap is formed. If the width of this region is x (see Fig. l), then the meniscus movement, y , corresponding to an increase of x from x = 0 to x = x, is given by- where D, is the apparent diameter of the meniscus.For large values of x and y, or when overlap occurs over more than the central portion of the meniscus, i.e., where D, is no longer sufficiently uniform, the value of y given by equation (1) is too small. Such errors do not normally arise until the meniscus has appeared below the bottom of the graduation. Values of D, were determined for water in ordinary glass tubing of various internal diameters (DJ. For D, Q 0.6 cm (which normally includes all sizes of pipette up to 50 d) a linear relationship was observed- AS Dt increased above 0.6 cm, D, increased rapidly until at Dt = 1.2 cm it was about twice as large as the value given by equation (2). Despite their subjective nature, three kinds of setting to the mark may be distinguished in terms of the value of x- (1) Meniscus just on the graduation line (x = x,).This is obtainable only under the best conditions with "pneumatic control'' of the meniscus level. A position where the meniscus was perfectly tangential to the top of the graduation was never observed; there was always a region where contact was uncertain. (2) Meniscus moderately on the graduation line (x = x,). y = x"4Dm . . . . . . . . * * (1) D, T== 1*80Dt . . . . .. . . - * (2) This represents careful working with finger control of the meniscus level.436 DEAN AND HERRINGSHAW SOME ASPECTS [Vol. 86 (3) Meniscus generously on the graduation line, which is usually equivalent to very slightly below it (x = x?). Provided that the meniscus is viewed against a suitable background this setting should be the extreme case.If D, is very small, the meniscus cannot be distinguished accurately without a lens, whereas if D, is too large, the value of x depends too much on individual marks. For values of Dt between 0-1 and 0.6 cm, the following relationships were observed- This represents routine working. . . - * (3) x1 0.33 D,; x2 21 0.51 D,; x3 N- 0.64 D, . . From equations (l), (2) and (3), y and the corresponding volume (AV) between y = 0 and y = y can be expressed in terms of Dt- Setting .. .. 1 2 3* y, cm . . . . 0.015 D, 0.036 Dt 0.057 D, AV, ml .. . . 0.012 IDt* 0.028 Dt3 0.044 D,B * For setting 3, the calculated values of y are usually greater than the thickness of the graduation line and hence are in agreement with the definition of setting 3.However, a definition based entirely upon an emergence of the meniscus below the graduation line is unsatisfactory, because (a) graduation lines vary in thickness, (b) it is difficult to tell when the meniscus begins to appear below the line and (c) there is considerable refraction around the edges of the mark, which makes the apparent thickness of the mark, as measwed by meniscus movement, greater than the true thickness. Other things being equal, the volume AV represents the range within which the volume contained by the pipette at the given setting should lie. The corresponding standard deviation (S,) of this contained volume cannot be calculated accurately, but is likely to be about AV/2 (see Conway3). The values in 'Table I are typical and are used later in the calculation of the over-all precision of the volume delivered.Pipette volume, ml M2 5 10 10 20 50 M50 TABLIZ I TYPICAL VALUES Setting I),, cm x < 0-2 mm 0.1 3 0.37 2 0.36 1 0.36 1 0.47 1 0.62 x < 0.2 mm 0.10 AT', ml 0-0004 0-0023 0.0013 0.0006 0.0013 0.0029 0.0004 0-0002 0.0012 0.0006 0.0003 0.0006 0.0016 0.0002 TABLE I1 RESULTS FOR GRAIIE-B BURETTES Burette Delivery Number of volume, Setting time, Dt, A V , A V / 2 , S,/ .\/2, duplicates ml seconds cm ml ml ml 6 2 210 0.326 0.00097 0.0005 0.0004 17 50 -* 300 1.1 1 0.0082 0-0041 0.0030 16 * This was slightly better than setting 2. Because Dt is large, the value of AV was calculated from the observed values of x, D, and Dt. For burettes with suitably long delivery times, errors other than those in reading the position of the meniscus are negligible, and a direct comparison between the calculated and observed precision is immediately possible.The results in Table I1 were obtained with grade-B burettes fitted with a reading device to avoid parallax and with all readings made at a mark; the estimate of the standard deviation in reading the burette is AV/2 as before, and this should be equal to 1/43 times the observed precision (S,) of the volume delivered. The observed values of S, were found to be independent of the volume delivered. The agreement between AV/2 and S,/dz is remarkably good in view of the approximations involved. Manipulation of pipettes when setting to the mark--It is customary to adjust the meniscus to the mark with the tip in contact either with the liquid surface or with a wetted surface.When contact is broken there is an inevitable change in the associated surface-tension forces,July, 19611 OF PIPETTE USAGE. PART I. PRECISION 437 and this, by virtue of the elasticity of the air enclosed between the mark and the end of the pipette, causes the liquid to retract slightly. The most reproducible conditions are obtained by holding the tip against a clean wetted surface at a large enclosed angle. If the surface is neither clean nor wet the error can be appreciable, e.g.- Volume delivered by 10-ml pipette at 20" C- Tip held against clean wetted surface-9.9973 f 0.0008 ml (12) Tip held against dirty surface-9.9996 0.0019 ml (5). The elasticity of the enclosed air also requires that the filled pipette be handled gently to avoid expulsion of the liquid by jolting; similarly, the tip should never be wiped with tissue at this stage (compare Conway3) otherwise an appreciable volume of liquid may be removed.ERRORS ASSOCIATED WITH THE TIP- Owing to capillary forces a small volume of liquid (the tip-volume) is always left in the tip of the pipette after delivery has ceased. This is of no consequence provided that the tip-volume remains constant from delivery to delivery. At equilibrium, the surface-tension forces inside the tip are balanced by the surface- tension forces outside the tip P h s the hydrostatic force of the liquid column inside the tip. If the attitude of the tip and the density and surface tension of the liquid are constant, then the reproducibility of the tip-volume is determined solely by the reproducibility of the surface-tension forces outside the tip, i.e., those forces arising from the curvature of the liquid surface around the outside of the tip.This curvature is markedly affected by the mode of contact between the tip and the liquid; thus it is dependent upon whether the tip,is TABLE I11 EFFECT OF CHANGE IN SURFACE TENSION ON TIP-VOLUME Pipette volume, ml . . .. .. 10 20 60 Tip-volume with water . . .. .. 0.0133 0.0203 0.0178 Tip-volume with Teepol solution . . 0*0070 0*0090 0.0075 ( y = 73 dynes per cm), ml . . . . ~0.0003(4) &0-0008(4) 0*0005(7) ( y = 30 dynes per cm), ml .. . . +0*0005(3) +0*0006(3) rtl: O.OOOS(8) held on the bulk liquid surface or on a wetted surface, and, if the latter, upon what angle the tip makes with the surface.The effect in terms of a change in tip-volume is shown by experiments with the 10-ml pipette held vertically. The differences between the tip-volume (determined by weighing the liquid expelled) obtained after withdrawing the tip from a position just touching the liquid surface and that after withdrawing the tip from different positions were- Touching wetted glass wall, enclosed angle 60" . . +0.0015 ml Touching wetted glass wall, enclosed angle nearly 0" +Om003 ml Immersed 1 mm in the liquid, hurried withdrawal . . -0.003 ml (Total tip-volume: 0.013 ml; internal diameter of tip at meniscus: 2 mm, at end of tip: 0.7 mm.) The effect of a change in surface tension (7) on the tip-volume (obtained after withdrawing the vertically held pipette from a position just touching the liquid surface) together with typical values for the tip-volume are shown in Table 111.The reproducibility of the tip- volume is seen to be independent of tip-volume, the average value being +0-0005; this has to be corrected for the reproducibility of the minute volume of liquid left after expulsion (but which cannot be determined accurately) and the corrected reproducibility is estimated to be about +0-0004 or less. The error caused by variability of the tip-volume is thus much less than the possible variation in the tip-volume itself. Although a reduction of the internal diameter at the tip serves to diminish some of the errors so far considered, in principle the only way to eliminate them is to expel the tip-volume and include it in the volume delivered.Most authorities (without apparent justification) condemn such a procedure; admittedly it is impossible in practice to expel all the liquid in the tip, and with the above pipettes the amount of liquid remaining was about 04004m1, although up to 0.0008 ml was occasionally observed. Complete expulsion is hindered by the large amount of glass at the tip (as indicated in a British Stan- dard*); if the tip is drawn out to leave a minimum of glass at the tip, as in the modified pipettes, the volume remaining after expulsion can be reduced to about 0.0002 ml. Even so, this is very small.438 DEAN AND HERRINGSHAW : SOME ASPECTS [Ivol. 86 Considering the effect of surface tension alone on the tip-volume, as intrinsic changes greater than 2 dynes per cm are unlikely with dilute aqueous solutions, there is no need to expel the drop, provided that the treatment of the tip at the end of the delivery is rigidly standardised.If this is not done, or cannot always be done, a procedure involving expulsion of the tip-volume may be the best for varied use. VARIATION OF THE DELIVERY TIME AND ASSOlCIATED ERRORS- Owing to the slowness of flow at the end of the delivery it is difficult to determine with any precision the time a t which flow stops. For this reason, deliveries were timed to a mark about 3 cm above the tip, the meniscus movement being sufficiently rapid at this mark to allow accurate timing. For vertically held pipettes the delivery times (in seconds), measured by a stop-watch reading to 0.1 second, are shown in Table IV.TABLls Iv DELIVERY TIMES OF PIPETTES Temperature: 16" C 10-ml pipette 60-ml pipette seconds seconds seconds seconds 1 -- ' Full delivery, * Lower stem, Free delivery . . . . 18.65 & 0.07(6) 4.60 & O.OO(3) 41.63 & 0.07(11) 3.30 & O.OO(3) Tip immersed 0.5 cm in water . . .. . . 19.10 0*07(6) 4-60 & 0.05(4) 42.36 & O.OO(3) 3.35 & 0.05(5) Tip held against wetted surface .. .. 18.07 & 0-07(6) 4.25 & 0*06(5) 41.25 0*05(3) 3.10 0*10(6) The differences in the delivery times caused by a change in the method of emptying the pipette are as expected. The differences between the times of delivery from. the lower stem are smaller than those for the whole delivery, indicating that a change of conditions affects the whole delivery rather than that from the lower stem alone.A change in delivery time causes a change in the volume of liquid remaining on the walls after delivery (the wall-film). As shown by Conways and confirmed by us (see Part 11), the wall-film is proportional to (delivery time)--*; therefore, from the values for the wall-film given in Part 11, a change of 0-5 second in the delivery time produces a change of 0.0006 ml in the wall-film (and consequently, by difference, in the volume delivered) of, fortuitously, either pipette, other things being equal. Determinations of the time for full delivery to the tip showed a precision of +0.15 second, which is a combination of personal error and a dis- continuous time distribution ; the real variation is less than k 0.1 second (compare Conways), which, with constant delivery technique, corresponds to a precision for the wall-film of better than +O.OOOl ml.TABLE v OVER-ALL PRECISION 0 1 7 VOLUME DELIVERED Pipette volume, ml M2* 5 10 20 50 M50 Remarks Setting N0.t x <Om2 mm 3 2 1 1 1 x<0*2mm Delivery time, - -7 seconds .. 22 15 22 22 31 44 41 - s, . . .. . . 0~0000 - 0.0001 0.0001 0.0002 0.0004 0.0004 = 0.06" C s, . . ,. . . 0~0000 - 0.0002 0.0002 0.0002 0-0002 0.0002 = Tf: 2seconds S, (observed) . . 0.00028 0.0016 0*0008 0.00049 0.0008 0.0016 0-0006 Pipette clamped No. of determina- S, as percentage of s, . . .. . . 0*0002 0.0012 0-0006 0.0003 0.0006 0.0015 0.0002 - s, . . .. . . 0.0002 - 0.0004 0*0004 0.0004 0.0004 0.0002 - s, .. .. . . 0-0000 - 0~0001 0~0001 0*0001 0~0001 0~0001 - S t (calculated) . . 0.0003 0.0014 0.0008 0.00056 0.0008 0.0016 0.0006 - (See Part 11) vertically - tions. . .. 11 10 17 + 12 15 9 12 50 total delivery 0.02 0.03 0.008 0.005 0.004 0-003 0.001 - * Drainage time, 6 seconds. t See p. 435.July, 19611 OF PIPETTE USAGE. PART I. PRECISION 439 Unfortunately, a change in the wall-film due to slight vibrations imparted by the hand (see Part 11) is not reflected by a corresponding change in the delivery time- 50-ml pipette, clamped vertically . . . . 40.90 3. 0.00 (4) 50-ml pipette, hand-held vertically . . . . 40.90 0.05 (3) (The difference between the above and previous values for the same pipette is due to the use of a different mark.) Delivery time, seconds Further, there are other factors (e.g., surface tension and viscosity-see Part 11) that can affect the wall-film with but little effect upon the delivery time.Consequently, any con- clusions regarding the wall-film that are based upon delivery times must be made under conditions in which such influences are excluded. The direct determination of the reproducibility of the wall-film is subject to so many errors that it was not attempted. OVER-ALL PRECISION OF THE VOLUME DELIVERED- expressed as a volume), i.e.- This will be the resultant of the component reproducibilities (Sn, standard deviations Setting to the mark Temperature uncertainty s1 S, } volume before delivery. Precision of the total Tip-volume Wall-film Drainage 2 } Precision of the residual volume after delivery. s, As all these components are independent, the over-all variance ( S t ) will be the sum of the individual variances, i.e., SF = XS,,. The individual variances have been measured or estimated for a number of pipettes, and the results, together with the observed values for the over-all reproducibility, are given in Table V. The agreement between the observed and calculated values of St is remarkably good and confirms the general correctness of the appraisal of the component errors. The largest error is in setting the meniscus to the mark. However, incorrect or unsuitable technique can increase the other errors by a factor of ten or more (see Part II), whereas the value of S, would remain sensibly constant. The experimental values for St are in agreement with the best of the somewhat frag- mentary data available in the l i t e r a t ~ r e . ~ ~ ~ ~ ~ They are significantly better than those found by Conway,3 and this must be ascribed to the technique he employed. REFERENCES 1. 2. 3. 4. British Standard 1583:1950. 6. 6. 7. Bishop, E., Anal. Claim. Ada, 1959, 20, 315. Strouts, C. R. N., Gilfillan, J. H., and Wilson, H. N., Editovs, “Analytical Chemistry: The Working Conway, E. J., “Microdiffusion Analysis and Volumetric Error,” Fourth Edition, Crosby Lockwood Ponndorf, W., 2. anal. Chem., 1931, 84, 289. Stas, J. C., Ann. Chim. Phys., 1872, 25, 89. Hoitsema, C., 2. angew. Chem., 1904, 17, 648. Tools,” Oxford University Press, 1955, Volume I, p. 79. and Son Ltd., London, 1957. Received January 24th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9618600434

出版商:RSC    

年代:1961

数据来源: RSC

摘要:

440 DEAN AND HERRINGSHAW: SOME ASPECTS OF PIPETTE [Vol. 86 Some Aspects of Pipette Usage Y4art 11. The Effects of Certain Variables BY G. A. DEAN* AND J. F. HERRINGSHAW (Chemistry Depavtment, Imperial College of Science and Technology, London, S. W.7) Investigations have been made of the effects of surface tension, viscosity, the temperature rise and the vibrations caused by holding the pipette by hand, inclination of the pipette, thermal hysteresis and grease on the volume delivered by the pipette and on 0th-er relevant properties of the pipette. Conway's relationship between the delivery time and the wall-film is con- firmed, and his work on the effect of temperature on the wall-film is corrected and extended. The most serious sources of variable error are summarised and recommendations are made whereby these can be minimised without undue inconvenience. THE variables considered can be divided into two groups: (a) those over which little or no control is possible, e.g., changes in surface tension, and (b) those that can be controlled, most of which are variations in technique. Investigations of the first group have been made to discover their modes of action and the magnitudes of their effects upon the volume delivered.Investigations of the second group are similar, but, in addition, conditions have been determined in which their effects on the volume delivered are minimised. IXFLUENCE OF SURFACE TENSION- per cm) or with a 1 per cent. v/v solution of Teepol (y = 30 density 1.000932 2 0.000017 (5 + 5)). Deliveries were made from vertically clamped pipettes filled with water (y = 73 dynes 1.5 dynes per cm; relative The results are shown in Table I.TABLE I EFFECT OF SURFACE TENSION 10-ml pipette 20-ml pipette M60-ml pipette , 2oo Jwater, ml 9.9973 0-OOOS(1.i) 20.0221 3- O.OOOS(9) 48.4950 & 0.0006(5) \Teepol, ml 9.9841 0.0002(5) 20.0133 0*0007(4) 48.4563 & O.OOOS(4) Difference, ml . . . . . . 0.0132 0-0088 0.0387 Volume of wall- water, nil 0.050 0.063 0.084 film* { Teepol, ml 0.067 0.084 0.121 *Obtained by subtracting the volume delivered plus the tip-volume from t h e total capacity and The presence of Teepol increases the wall-film by about 40 per cent., and, although this increase is partly compensated by the reduction in the tip-volume (see Part I), the volume delivered is less.Decrease of surface tension also decreases the meniscus volumel; for a tube of diameter 0.5 cm, a change in surface tension from 73 to 40 dynes per cm reduces this volume by 0-001 ml, The change in meniscus volume decreases rapidly as the diameter of the tube is decreased; hence the effect here is small. The presence of Teepol also affects the delivery time; for the 10- and 50-ml pipettes the delivery times were decreased by about 0.6 second. Such a change corresponds to an increase of about 0.0006 ml in the wall-film of, fortuitously, either pipette. Alternatively, if the decrease in delivery time is attributed tcl a decrease in the viscosity (see later), this corresponds to a decrease of about 0.006 ml in the wall-film. Clearly, the major change in the wall-film is not caused by a change in the delivery time.If the change in the wall-film with surface tension is linear, then a change of 1 dyne per cm in surface tension produces a change of about 1 per cent. in the wall-film; as the differences in surface tension encountered with dilute aqueous solutions are usually not greater than 2 dynes per cm, the errors here are not serious. Errors due to the decrease in surface tension by contamination may be more significant, but these can usually be avoided by taking suitable precauti0ns.l correcting to 20" C (see later). Whatever the conditions, a small wall-film is clearly advantageous. * Present address : Chemistry Department, Dalhoiisie University, Halifax, Nova Scotia, Canada.July, 19611 USAGE. PART 11. THE EFFECTS OF CERTAIN VARIABLES 441 INFLUENCE OF VISCOSITY- A change in viscosity may result from the presence of some substance in solution or from a change in temperature; the former is considered below and the latter is considered on p.445. The viscosity of water was increased by the addition of a boiled and filtered solution of a cellulose wallpaper-adhesive, chosen for its negligible effect on the surface tension and for the small concentration required to change the viscosity markedly. The resulting viscosity was determined by treating the modified 50-ml pipette as a Redwood viscometer, when the following equation is applicable- where A and B are constants and t is the delivery time in seconds. B (1.20) were determined by calibrating with water at different temperatures.in Table 11, y / d = At - B/t is the viscosity in poises, d is the density of the solution in grams per millilitre, The values of A (0.00097) and The effect of a change in viscosity on the properties of the M50-ml pipette is shown TABLE I1 Temperature: 18.5" C INFLUENCE OF VISCOSITY Solution Water 1 2 3 4 Relative density . . . . 1~00000 1.00016 1.00035 Specific viscosity . . . . 1.00 1.5 2-1 Surface tension, dynes per cm 73 72 72 Drainage,* ml .. . . 0.001 0.001 0.001 0.141 0.180 Delivery time, seconds . . 41 44 48 Volume, D 20" C, ml . . 48-4958 48.4403 48.401 1 Wall-film, ml . . . . . . 0-085 & 0.0005(5) 0*0007(4) 0-0023(4) * Over the standard drainage time of 22 seconds. 1.00070 3.2 72 66 0.002 48.3567 0.224 - +0*0009(4) f 1-00 135 5.1 72 72 0-002 48.2957 10.0020(4) 0.285.An increase in the viscosity causes a decrease in the volume delivered. This cannot be the result of increased delivery time, because the effect is too large and in the wrong direction. If the volume of the wall-film is plotted against viscosity, a shallow curve passing through the origin is obtained. From this, a change of t 1 per cent. in the viscosity of water produces a change of about 20.8 per cent. in the wall-film; such a change in viscosity is a typical value for 0.1 M solutions. As with surface-tension effects, a small wall-film is advantageous. EFFECTS ASSOCIATED WITH HOLDING THE PIPETTE BY HAND THERMAL EFFECTS- If the pipette is held by hand, appreciable heat may be imparted to it in a remarkably short time. The results in Table 111 were obtained with a 10-ml pipette held at the extreme top, delivery being made on to the bulb of a thermometer.The figures quoted are the differences between the temperature (in O C) recorded on the thermometer during the delivery and that of the water in the reservoir. (The difference in temperature between the air and the water in the reservoir never exceeded 1" C). TABLE I11 THERMAL EFFECT OF HOLDING 10-ml PIPETTE BY HAND Held by Held by hand gloved hand 1. Held at extreme top .. . . . . . . .. . . 0.1 not detected 2. Bulb held through delivery . . .. . . .. . . 1.5 0.2 3. Lower stem held throughout delivery . . . . .. . . 0.7 0.2 4. Lower stem held for adjustment only . . .. .. . . 0-5 not detected 5. As 4, rapid adjustment . . . . .. .. .. . . 0.2 not detected 6.Lower stem held by plastic-foam cube throughout delivery . . not detected - These differences are quite significant; a change of 0.5" C in the temperature of the water could in certain circumstances lead to an error of 0.001 ml, which is larger than the442 DEAN AND HERRINGSHAW: SOME ASPECTS OF PIPETTE [Vol. 86 observed precision of delivery for this pipette. Thus in precise work handling the pipette without adequate insulation should be avoided. The conversion of kinetic energy into heat can affect the pipette only when the liquid is sucked into the pipette. The temperature rise can be shown to depend upon the speed of filling, but for normal pipettes is not likely to exceed 0.02" C. VIBRATION EFFECTS- These can be demonstrated with a 50-in1 pipette and a strongly coloured solution; the thickness of the wall-film in the bzdb (as indicated by the depth of colour) is greater when the pipette is held by hand during delivery than when it is clamped in position. This can only be caused by the slight vibrations imparted to the pipette when it is held by hand; these vibrations produce oscillation of the meniscus level, which simulates an apparent increase in the rate of descent and thus leads to a larger wall-film.In the experiments (see Table IV) carried out to investigate this effect, the hand-held pipette was supported with one hand at the extreme top and the other at the foam insulator, both elbows being on the bench; this position was slightly less steady than when the pipette was held at the extreme top only and with the elbow on the bench.Absolute steadiness could not be obtained, even with the forearm resting on a solid support. TABLE IV THE EFFECT ON DELIVERY OF HOLDING THE PIPETTE BY HAND Internal Pipette diameter volume, of bulb, ml mm 10 15 20 20 50 29 M50 29 Volume, D 20" C Clamped, Hand-held, ml ml 9.9973 rfi 0-0008(17) 9.9973 & O*OOOS(lZ) 20.0221 & 0*0008(9) 20.0194 & 0.0013(10) 50.0641 & 0*0016(12) 50.0493 -f 0-0034(16) 48.4950 0.0006(5) 48-4804 & 0-0045(6) 48.4563 & 0-OOOS(4) * 48.4502 0.0020(4) * Wall-film (with pipette Difference, clamped), ml ml 0~0000 0.050 0.0027 0.063 0.0148 0.108 0.0146 0,084 0.0061* 0*121* * Results with Teepol solution; y = 30 dynes per cm. The effect of vibration is more pronounced as the capacity increases, and this is probably related to the increasing diameter of the bulb.The thickness of the wall-film in the bulb of the 10-ml pipette is greater than that of the 50-ml pipette (0.0023 and 0.0011 cm, respec- tively) and the thicker wall-film may accommodate the movement better. This view is corroborated by the decreased effect observed with a solution of Teepol; the wall-film is increased and the effect of vibration is more than halved. The effect here of surface tension could not be determined. If the effect of vibration is indeed dependent on the thickness of the wall-film, then, other things being equal, it should increase as the delivery time is increased, thereby obviating any other advantages to be gained by this procedure. The effect of extreme vibration on the volume delivered by the 10-ml pipette is shown in Table V.TABLE v EFFECT OF VIBRATION ON VOLUME DELIVERED FROM 10-ml PIPETTE Condition Volume, D 20" C, ml Difference from volume in 1, ml 1. Clamped . . . . . . 9.9973 ~t 0*0008(17) - 2. Gentle vibration . . . . 9.9831 rl: 0.0017(5) 0.0141 3. Vigorous shaking . . . . 9.9593 ~b 0*0124(5) 0.0380 In agreement with Ponndorf2 and Bishop,3 but not C o n ~ a y , ~ we find that manual holding can significantly affect the precision of the volume delivered, although the effect is small compared with the change in the volume delivered. No increase in the wall-film was observed with the 5- and 50-1111 burettes (Part I) even with gentle vibration or knocking. INFLUENCE OF INCLINATION OF A PIPETTE ON THE DELIVERY AND DRAINAGE- A change in the inclination of the pipette has three distinct effects; first, the delivery time is altered because of the change in the head of liquid, secondly, the drainage is affected a,nd thirdly there is a change in the circumference of the meniscus over those parts of theJuly, 19611 USAGE. PART 11.THE EFFECTS OF CERTAIN VARIABLES 443 pipette, e g . , the bulb, where capillary forces do not predominate. The net result of these changes depends largely upon the individual characteristics of the pipette. If the wall-film is not too large and the pipette is vertical, drainage from the upper stem and the bulb travels only slowly down the pipette and does not reach the tip until the normal drainage time of 15 to 20 seconds has elapsed. Thus the only drainage from which errors may arise is that from the lower stem.If the wall-film is large, as after rapid deliveries, the rate of drainage is increased and the above is no longer true. If the pipette is tilted, drainage occurs in two main sections down each side of the pipette. The liquid from these two sections collects along the lower wall of the pipette and a relatively large volume of liquid accumulates there. Most of this runs rapidly down the length of the pipette and reaches the tip usually before the normal drainage time of 15 to 20 seconds has expired. This will occur whether the pipette has been emptied in a tilted or in a vertical position. This view is confirmed by the results shown in Table VI, obtained with the M50-ml pipette clamped in position; the drainage was measured by observing the liquid level in the calibrated tip .TABLE VI EFFECT OF INCLINATION ON DRAINAGE FROM M50-ml PIPETTE Angle to vertical . . .. . . . . .. .. 0" 30" 60" Delivery-time, seconds . . . . .. . . . . 41.6 47 60 0 to 20 seconds, ml . . .. * . 0.001 0.009 0-017 { 20 t o 40 seconds, ml . . .. . . 0-013 0.016 0.014 Rate of drainage, ml per 20 seconds a t 20 seconds . . 0-003 0.012 0.017 Provided that the minimum possible drainage time was used, a change in inclination produced little or no change in the volumes delivered by our pipettes. With drainage times of 20 seconds, an increase (AV) in volume delivered was observed due to increased drainage. The values of AV (see Table VII) agree well with the independently determinedincreases in drainage.Pipette, ml . . Angle to vertical 2 seconds' 20 seconds' TABLE VII VALUES OF AV . . . . . . .. .. .. 10 M50 - .. . . .. .. .. .. 45" 30" 60" drainage, ml . . .. .. . . 0.009 0.01 2 0.0 15 - - drainge, ml . . . . .. . . 0.000 Increise in total drainage, 0 to 20 seconds, ml . . . . 0.010 0.008 0.016 These results in Table VII suggest that the effect of the increase in delivery time obtained on tilting the pipette is a subsidiary effect and is compensated by the effect of the change in the circumstances of the meniscus, the exactness of the compensation depending probably on the geometry and the delivery time of the pipette. It follows from the foregoing that no general conclusions can be made about the effect of inclination of the pipette.However, it is clear that, for consistent delivery and minimum drainage, the best position for the pipette is a vertical one-this is also the most well defined. It is also advantageous to use a short (a few seconds) drainage time, since the rate of drainage increases rapidly during the first minute; on the other hand, owing to the slowness of flow towards the end of the delivery, the drainage time must not be too short, otherwise errors may arise from a premature withdrawal of the pipette. It is interesting to note that the National Physical Laboratory proposes to alter its standard drainage time from 15 to 3 seconds, although, according to an initial report,6 this is in the interests of speed rather than from considerations such as those described above. There is also a smaller absolute error in counting 3 seconds than in counting 15 seconds.INFLUENCE OF DELIVERY TIME ON THE WALL-FILM- Other things being equal, a longer delivery time (t) results in a smaller wall-film; according to ConwayJ4 the wall-film is proportional to t-* both for pipettes and burettes. This relationship was confirmed as described below.444 DEAN AND HERRINGSHAW: SOME ASPECTS OF PIPETTE [Vol. 86 The delivery times of the lo-, 20- and M50-ml pipettes were varied by partly closing the top of the pipette with a rubber tube and a screw-clip.* The volumes of water and of Teepol solution (y = 30 dynes per cm) delivered for different delivery times were determined with the pipettes clamped vertically and with standard drainage time (22 seconds). 2 (Delivery time, seconds)-ix 10' Fig.1. Influence of delivery time on the wall-film; results for the lo-, 20- and M5O-ml pipettes: a, with water; 0, with Teepol solution After the appropriate tip-volume had been added, these volumes were plotted against 19, when a linear relationship was obtained (see Fig. 1). The deviations that occur for t > 6 minutes are probably due to evaporation or to very slow final emptying (compare Stott6). TABLE 'VIII CALCULATED AND DETERMINED VALUES FOR WALL-FILM Pipette, ml . . . . . .. 10 20 50 M50 Wall-film,* ml . . Wall-film by direct determination, ml the total capacity. . . 10.060 20.105 - 48.581 48.581 . . .. . . 0.060 0.063 - 0.084 0-063 0.108 0-082 * Calculated by subtracting the values for the volume delivered and the tip-volume from If the lines are extrapolated to t-* = 0, the total capacities of the pipettes can be obtained with reasonable accuracy.By subtracting froin these values the volume delivered under normal conditions and the tip-volume, the wall-film for normal delivery conditions can be obtained. Some results are shown in Table VIII. The values for the wall-film are corrected His experi- ments suggest that, with the same pipette, the two methods give different results, but when his results are re-plotted as a relationship between volume delivered and t-B, parallel straight lines are obtained. If these are extrapolated to t-& = 0, different total capacities are indicated. This is probably caused by a change in the tip-volume, which was not determined. Total capacity : : .. 10.059 20.106 - i 0-051 * This is preferable to the method used by Stotk6 La., grinding off small portions of the tip.July, 19611 USAGE. PART 11. THE EFFECTS OF CERTAIN VARIABLES 445 to 20" C (see later). The direct determinations were made by filling the pipette to the mark with 0.02 M potassium permanganate, care being taken not to rinse above the mark. The pipette was clamped vertically, and emptied and drained in the standard manner; the tip- volume was then expelled. The liquid remaining in the pipette was rinsed into a 50-ml calibrated flask, the solution diluted to the mark, and the permanganate determined absorptio- metrically. The total capacity obtained is the same whether water or Teepol solution is used and the calculated values for the wall-film agree with those determined directly.The results therefore confirm the linear relationship between the volume of the wall-film and t-4 for pipettes clamped in a vertical position. In this report, all corrections to volumes delivered for changes in delivery time have been calculated on this basis. For hand-held pipettes, a similar relationship applie~.~ .48* 5 - =I E $ 48.50- 0 cl n g 45.49- 8 - 48.48- I - / - 0.09 -*o- I0 a z -0.11 : -0.12 -F 3 0 5 ,L 10 15 20 25 30 35 ; 0.13 ' INFLUENCE OF TEMPERATURE OTHER THAN EXPANSION EFFECTS- The corrections for temperature changes normally applied are those incorporated in tables for the effect of temperature on the density of water, on the capacity of the glass vessel and on the density of air (calibration only). As both viscosity and surface tension are tem- perature-dependent, the wall-film should be affected by changes in these quantities with temperature.Conway4 has determined the effect for a 25-ml pipette (see Fig. 2) and his results are discussed subsequently. The effect was measured by determining the volume delivered by the M50-ml pipette (clamped vertically) at various temperatures. These volumes were then corrected to 20" C in the normal way. When these volumes are plotted against the temperature of calibration, they lie on a straight line of slope 0.0012 ml per "C (see Fig. 2). This change can be related to the wall-film as shown below- This work Conway Change in wall-film, ml per "C . . - 0.0012 - 0.00 16 Wall-film a t 20" C, ml.. . . . . 0.084 0.110 Change of wall-film, % per OC .. -1.4 2 0.1 -1.45 0.1 The figure given by C~nway,~ and quoted by Sutton,' for this variation with temperature is 0-006 per cent. of the whole delivery for an increase of 1" C. This figure is of little value, as it applies only to the particular pipette that Conway used. The change must be related to the wall-film, and the figure - 1.4 per cent. of the wall-film per "C is of general applicability. The observed effect of temperature on the volume delivered at 20" C by the M50-ml446 DEAN AND HERRINGSHAW: SOME ASPECTS OF PIPETTE [Vol. 86 pipette can be compared with that calculated from the separately estimated changes due to viscosity and surface tension. If the temperature is increased from 15" to 25" C-- .. 0.012ml Observed increase in volume delivered .. .. . . .. Correction for decrease in delivery time (1.45 seconds) . . .. . . +0.002 ml Correction for decrease in surface tension (1.5 dynes per cm) . . . . +0402 ml Corrected increase, which should be due to viscosity change alone . . 0.016 ml The decrease in viscosity is 25 per cent. referred to that at 20" C. Hence the decrease in the wall-film, from p. 441, is 0.017 ml. These calculations show that the main contribution to this influence of temperature is the effect of a change in the viscosity. As the observed effect is directly proportional to the wall-film, a small wall-film is again advantageous. For accurate work, the effect is sufficiently large to require an additional correction to be made. DISTRIBUTION OF THE WALL-FILM- Conway4 has shown that the volume of the wall-film of a vertical tube can be related to the diameter (d in cm) and length (I in cm) of the tube, and to the rate of descent of the meniscus (v in cm per sec,), thus- Wall-film, ml = 0.00345 d 1 e 5 ZuO.5 (the temperature at which this equation is valid is not stated).stem of the 10- and 50-ml pipettes (see Table IX). The agreement is satisfactory. This was used to calculate the wall-films in the upper stem, the bulb and the lower Length,* cm .. .. . . Time of meniscus descent, seconds Internal diameter, cm . . . . Wall-film (calculated), ml . . Percentage of total wall-film . . Total wall-film (calculated), 11-11 . . Total wall-film (measured), correc- ted to 16" C, ml . . .. TABLE IX WALL-FILM DISTRIBUTION Temperature : 16" C 10-ml pipette Y h - - - - - - 7 TJpper Lower stem Bulb stem 5-0 6.5 17.5 0.9 13.3 7.0 0.36 1.4 0-36 0.018 0-018 0-018 33 33 33 -Y 0.055 0.053 50-ml pipette -7 Lower Upper stem Bulb stem 10.0 8.5 18.0 2.1 36.6 5.0 0.62 2.8 0.43 0.036 0.044 0.033 32 39 30 0-113 L .I Y 0-1 14 * Corrected for the change in diameter a t the shoulders. Despite the errors of measurement and the absence of a reference temperature, the agreement between the calculated and measured values for the total wall-film may be taken as confirmation of the essential correctness of Conway's equation. The results in Table IX show that the wall-film is about equally distributed between the three portions of the pipette.Useful reductions in the wall-film, and hence increases in accuracy, can therefore be obtained by reducing the dimensions of the stems a s much as possible.Thus for our 50-ml pipette, if the upper stem were reduced to zero length and the lower stem from 18 to 10 cm in length and from 0.43 to 0.20 cm in diameter, the wall-films on these portions would be reduced to nil and 0.012 ml, respectively, and the total wall-film from 0.113 to 0.057 ml. If, in addition, the delivery time were increased from 44 to 60 seconds, the wall-film would be further reduced to 0.049 ml, Le., an over-all reduction of about 60 per cent. The errors in the volume delivered due to changes in viscosity, surfa,ce tension and temperature would be correspondingly reduced. INFLUENCE OF THERMAL HYSTERESIS- The thermal hysteresis of volumetric glassware is usually assumed to be fairly large.Kolthoff and Stengers state that the effect is much smaller than has been previously supposed, but quote no data. After heating the M50-ml pipette (made from "soda" glass) to 100" C for 1 hour, an increase of 0-003 ml in the volume delivered was observed and was unchanged after 4 days. The effect here is small.July, 19611 USAGE. PART 11. THE EFFECTS OF CERTAIN VARIABLES 447 INFLUENCE OF GREASE- Conflicting statements appear in the literature about the influence of grease; its presence has been said to in~rease,~ decreaselo and both to increase and decreasell the volume delivered. Our own experiments with the 11150-ml pipette were inconclusive, except that in the presence of grease a large decrease in precision was observed. CONCLUSIONS The results of the investigation are in agreement with most of Conway’s conclusions regarding the wall-film.However, the precision of volume delivered obtained by Conway was relatively poor and therefore tended to obscure many effects that have been treated here. Thus, his conclusions about tip-technique, drainage, meniscus adjustment and the effects of holding the pipette by hand, although no doubt true in his work, are otherwise erroneous. Further, he suggested that the precision of delivery was proportional to the wdl-film and hence to (delivery time)-&. The work described here shows that the precision cannot be predicted from considerations of the wall-film alone. The most serious sources of variable error with a typical pipette are, in order of magnitude- (1) the vibrations caused by handling (larger pipettes only); (2) the effect of temperature on the wall-film; (3) a varying inclination (long drainage times only) ; a varying tip-technique ; the thermal effect of handling; all of which can be minimised by taking suitable precautions. If this is not done, the variation in the volume delivered can easily be more than ten times that quoted in Part I, i.e., in round figures, the precision of delivery will be kO.01 to +O.OZ ml depending on the size of the pipette. Such a variation may well be acceptable in practice, although it must be stressed that the differences in the volumes delivered from day to day may be greater. Useful increases in precision and accuracy can be conveniently obtained by keeping the pipette vertical at all times, by clamping the larger pipettes during delivery, by handling the pipettes with insulating material, by applying corrections for the effect of temperature on the wall-film and by using a short drainage time.Much can be done by intelligent selec- tion of the pipette, attention being directed to its shape and finish, the quality of the glass and the position, regularity and thickness of the mark. It is important that the delivery time should be within the limits prescribed by the National Bureau of Standards12 (in our opinion the lower limit prescribed by the British Standards Institute13 is too small).* Provided that these conditions are observed, a reliable technique is that described below. Attach to the top of the pipette a suitable “pneumatic-control device” (devices that include a large volume of air are not recommended), and clamp the pipette in a vertical position.Fill by suction to about 2 cm above the mark, and wipe around the tip with tissue. Adjust the meniscus to the mark by allowing liquid to flow out with the tip held against a clean wet surface at a large included angle. During this operation, the meniscus must be viewed against a suitable background. Sufficient attention is seldom directed to this point, whereas due care is usually taken when viewing the meniscus in a burette. We use a small white card with a black lower portion, attached to the pipette at the mark; any of the standard methods of reading a meniscus in a burette will probably serve equally well, provided that they are consistently used.Allow the pipette to deliver freely, i.e., with the tip in air, in a vertical position. When flow has ceased, allow 3 seconds for drainage and then either touch the tip against the surface of the liquid or expel the tip-volume by operating the “pneumatic-control device” with the tip held against the wall of the vessel. The second method must be used if delivery has been made into a vessel that contains another liquid. Make corrections for the appropriate expansion effects of temperature and the effect of temperature on the wall-film. Corrections for the effect of atmospheric pressure are required only in calibration. If the above procedure is adopted, then the precision of delivery quoted in Part I can be realised. The accuracy will also be high, i.e., the mean corrected volume delivered will * Since this paper was prepared, B.S.1583 has been revised (1961). The short (3 second) drainage time has been adopted and the minimum delivery time for some pipettes has been increased.448 BETHEL AND BOND: IODIMETRIC METHODS [Vol. 86 show little variation with time. Thus, the 10-ml pipette that was used in this work was calibrated, after delivery as above, seven times at intervals over a period of 2 years; the average of the volumes delivered (corrected tQ 20” C) was 9-9965 ml and all the calibrations were within 0.002 ml of this value. It will be noted that the recommended method includes allowing the pipette to deliver freely, whereas the British Standards Institute13 stipulates delivery with the tip held against the wall of the receiving vessel. In our opinion, free delivery is better, because (a) the conditions at the tip are more reproducible and (b) it is easier to keep the pipette vertical with free delivery. The British Standards Institute also specifies a different method of treat- ing the tip, i.e., holding the tip against a wetted surface. This could possibly be as reproducible as our method, i f the included angle were rigidly standardised, but no specific angle is stipulated. We thank Imperial Chemical Industries Ltd. (Billingham Division) for awarding a Bursary to one of us (G.A.D.), during the tenure of which this work was carried out. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES National Physical Laboratory, “Notes on Applied Science No. 6. partment of Scientific and Industrial Research, London, 1954. Ponndorf, W., 2. anal. Chem., 1931, 84, 289. Bishop, E., Anal. Chim. Acta, 1959, 20, 315. Conway, E. J ., “Microdiff usion Analysis and Volumetric Error,” Fourth Edition, Crosby Lockwood Rands, J. B., Chem. and Ind., 1958, 39, 1256. Stott, V., Trans. SOC. Glass Technol., 1921, 5, 307. Grant, J ., “Sutton’s Volumetric -4nalysis,” Thirteenth Edition, Butterworths Publications Ltd., London, 1955, p. 85. International Critical Tables, Volume V, p. 10. Kolthoff, I. M., and Stenger, V. A., “Volumetric Analysis,” Second Revised Edition, Interscience Publishers Inc., New York, 1947, Volume 11, p. 4. Grant, J., “Clowes and Coleman‘s Quantitative Chemical Analysis,” Fifteenth Edition, J. and -4. Churchill Ltd., London, 1947, p. 110. Wagner, J., 2. flhys. Chem., 1899, 28, 193. National Bureau of Standards Circular C434, Washington, 1941. British Standard 1583: 1950. Volumetric Glassware,” De- and Son Ltd., London, 1957. Received January 24th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9618600440

出版商:RSC    

年代:1961

数据来源: RSC

摘要:

448 BETHEL AND BOND: IODIMETRIC METHODS Iodimetric Methods of .Assay for Penicillin [Vol. 86 in Fermented Medium BY M. M. BETHEL AND C. R. BOND (Imperial Chemical Industries Ltd., Pharmaceuticals Division, Penicillin Plant, Trafford Par,?, Manchester 17) Two iodimetric methods are described for determining penicillin in fermented broth. In the first method, most of the impurities are removed by solvent extraction and the effect of the remaining impurities on the absorption of iodine is rendered negligible by adjustment of pH; the standard error of a single determination is approximately 2 per cent. The second method requires use of penicillinase, but is more rapid than the first, as it is direct and does not involve solvent extraction. This method depends on preliminary iodination of the impurities, during which penicillin is un- affected; the standard error of a single determination is also approximately 2 per cent.VARIOUS methods have been devised for assaying penicillin in fermented broth,l to 8 but we have found none of these to be sufficiently accurate or otherwise entirely satisfactory for the dual purposes of process control and assessment of experimental fermentations on a largeJuly, 19611 OF ASSAY FOR PENICILLIN I N FERMENTED MEDIUM 449 scale. During our work, we found that iodimetric assayg310J1 with inactivation by alkali could be applied to fermentation samples if two conditions were fulfilled, vix.- (i) Most of the organic impurities were removed by solvent extraction. (ii) The uptake of iodine by the remaining impurities, which otherwise interfered with the blank determination, was decreased to a minimum by adjustment of pH.EXPERIMENTAL One difficulty in developing a method for assaying fermented medium is that the penicillin contents of natural samples are unknown, and it is impracticable to prepare a fermented medium having a known content of penicillin. The use of an unfermented corn- steep liquor - sugar medium, with the addition of known amounts of penicillin, is not entirely satisfactory, as such a medium differs in composition from fermented samples. The closest approach to naturally fermented samples of known penicillin content appeared to be by the use of a partly fermented medium in which the penicillin had been destroyed. This was accomplished by slightly acidifying and then setting side overnight or heating at 120” C; after neutralisation, the complete destruction of the penicillin was demonstrated by micro- biological plate assay.Tests by the proposed methods A and 12 gave apparent-penicillin contents of up to 1 and 0.5 per cent., respectively. These contents, however, were too small to be of significance and were ignored in subsequent work on the samples. Known amounts of twice recrystallised sodium benzylpenicillin or potassium phenoxy- methylpenicillin, indistinguishable in potency from the current British National Standards for penicillin by both iodimetric and microbiological assay, were added. These standard preparations have potencies of 1670 units per mg for sodium benzylpenicillin and 1695 units per mg for phenoxymethylpenicillin (1530 units per mg for the potassium salt).12 Although the term “unit” relates strictly to biological activity, it has been widely used as a convenient expression of penicillin content in relatively low concentrations, particularly in fermented medium, and we have therefore used the term in this paper in the sense that 1 mg of sodium benzylpenicillin is equal to 1670 units and 1 mg of potassium phenoxymethylpenicillin is equal to 1530 units.These two values are almost exactly equivalent to each other when allowance is made for the difference between the molecular weights of the two compounds. We found that the second condition mentioned above could be satisfactorily met by decreasing the pH of the blank to 3.6, the lowest value to which benzylpenicillin can be subjected for the time necessary to complete the blank determination without causing significant inactivation of the penicillin present.Thus, when a typical inactivated fermen- tation medium derived from corn-steep liquor and reducing sugars, devoid of penicillin, was extracted with butyl acetate at pH 2, most of the impurities remained in the aqueous phase ; extraction of the butyl acetate solution with phosphate buffer solution yielded a relatively pure aqueous extract. Suitable aliquots of this extract were made alkaline with sodium hydroxide, acidified with hydrochloric acid and iodinated ; the blank solutions were iodinated directly after adjustment of pH to different values between 3-0 and 6-4. The amount of 0.01 N iodine absorbed by the test solution was 1.9 ml, and the results for the blanks were- pH .. . . . . . . 6.4 5-0 4.0 3.6 3.0 0.01 N iodine absorbed, ml . . 3-0 2.1 2-1 1-9 1.9 At pH 3.6 and below, the amounts of iodine absorbed by test and blank solutions were identical, so that there is no apparent interference from impurities at these values. A pH of approximately 3.6 can be readily attained in practice under the proposed conditions of assay by adding 1 ml of 1.0 N acetic acid to 10 ml of 0.01 M phosphate buffer solution. The extent of inactivation of benzylpenicillin under these conditions was deter- mined by measuring the absorption of iodine from a phosphate buffer solution containing 20,000 units of benzylpenicillin (much more than is normally present in a test) after a suitable addition of 0.02 N iodine had been made and the solution had been set aside at room tem- perature (approximately 21” C) for different periods.The excess of iodine was determined by titration with sodium thiosulphate solution ; the results were- Time of standing, minutes . . .. .. 0 5 10 15 20 Titre of 0-01 N sodium thiosuphate, ml . . 48-15 48.10 47.95 47.9 47.9 As the time needed for a blank determination is less than 5 minutes, the error arising from inactivation of the penicillin in the blank is negligible.450 BETHEL AND BOND: IODIMETRIC METHODS [Vol. 86 The extraction of penicillin from fermented medium at pH 2 into solvent is generally carried out under ice-cold conditions, because of the instability of benzylpenicillin in acid solution, However, such cooling of solutions considerably lengthens the assay and is there- fore undesirable in a method intended for routine tests on large numbers of samples; extraction at room temperature was therefore considered.This appeared to be practicable, as transfer of penicillin to the solvent can be carried out so rapidly that the likelihood of significant inactivation during extraction seemed small, and, once in butyl acetate solution, the penicillin is sufficiently stable to withstand any normal delays in the subsequent assay; tests for stability indicated a maximum loss of 0.2 per cent. per hour. Experiments on the efficiency of the solvent extraction at room temperature were therefore carried out on inactivated fermented medium containing known added amounts of sodium benzylpenicillin.The mean potency of the samples was found from twenty-two determinations to be 1160 units per ml, compared with a theoretical value of 1166. This is equival.ent to 99-5 per cent. of the original activity (or at least to 98.5 per cent., allowing for the possible small blank of the medium used), i.e., solvent extraction can be satisfactorily carried out at room temperature. The iodine equivalents of sodium benzylpenicillin and potassium phenoxymethylpenicillin were determined in solution in 0-01 M phosphate buffer under the assay conditions used; 1 ml of 0.01 N iodine was found to be equivalent to 0.395 mg (660 units) of sodium benzyl- penicillin and to 0-418 mg (640 units) of potassium phenoxymethylpenicillin. Under these conditions, the iodine equivalents were not found to vary significantly with concentration, provided that not more than half of the iodine present was absorbed.13J4 We have included it in this paper because it is a highly satisfactory assay procedure that can be used if penicillinase is un- available and also because it may be useful when information is required about penicillinase- resistant and penicillinase-sensitive penicillins.The method finally developed is described below. METHOD A : DETERMINATION OF BENZYL- OR PHENOXYMETHYLPENICILLIN I N FERMENTED MEDIUM AFTER EXTRACTION AND ALKALINE HYDROLYSIS REAGENTS- 2 per cent. of potassium iodide. Iodine, 0.02 N-Freshly prepare this solution each day from 0.1 N iodine containing Phosphate bufer solution, 0.02 M (PH 74)--Prepare from the sodium salts.Sodium thiosulphate, 0.01 N-Accurately standardised. PROCEDURE- Filter the sample, accurately transfer (by pipette) 10 ml of the filtrate, or of a suitable dilution if the potency exceeds 4000 units per ml, to a separating funnel, and add approxi- mately 50ml of butyl acetate. Add 1-Om1 of 8 per cent. v/v sulphuric acid containing 4 per cent. of cetrimide, shake vigorously for 30 seconds, and allow to separate. Discard the lower aqueous layer, and wash the butyl acetate layer with 10 ml of water. Add 75 ml of the phosphate buffer solution, shake vigorously for 30 seconds, allow to separate, and run the lower aqueous layer into a 250-ml calibrated flask. Rinse the separating funnel with 50ml of the buffer solution, add the rinsings to the contents of the flask, and dilute to the mark with water.Transfer a 20-ml portion of this solution to a 150-ml stoppered flask, add 5ml of 1.0 N sodium hydroxide, and set aside for 5 minutes. Acidify with 5ml of 1.1 N hydrochloric acid, add 5 ml of 0-02 N iodine, and set aside for 5 minutes (exclude light during iodination). Titrate the excess of iodine with 0.01 N sodium thiosulphate added from a 10-ml burette graduated in 0.01-ml divisions; use 1 ml of a 1 per cent. solution of starch as indicator. To determine the blank value, transfer a 20-ml portion of the buffer solution to a flask, add 2 ml of 1.0 N acetic acid, and mix. Add 5 ml of 0.02 N iodine, and immediately titrate with 0.01 N sodium thiosulphate; add 1 ml of the starch solution near the end-point.If difficulty is experienced in ascertaining the end-point, carbon tetrachloride should be used as indicator. Calculate the potency of the sample from the equation- (B - A) x F x E x 250 Potency, units per ml = - Volume of sample, ml x 200 in which A and B are the volumes of 0.01 N sodium thiosulphate needed for the test andJuly, 19611 O F ASSAY FOR PENICILLIN I N FERMENTED MEDIUM 451 blank solutions, respectively, F is the factor of the sodium thiosulphate solution and E is 660 for benzylpenicillin or 640 for phenoxymethylpenicillin. ACCURACY OF METHOD- Determinations on samples prepared for experiments on the efficiency of extraction indicate that the method gives results close to the theoretical values; in fact, results are low by about 0-5 to 1.5 per cent.The reproducibility of the method is good, the standard error of a single determination being approximately 2 per cent. This is based on the results in Table I for replicate tests on several samples of fermented media. TABLE I POTENCIES FOUND FOR FERMENTED SAMPLES IN REPLICATE DETERMINATIONS Tests Nos. 1 and 2 were carried out by one operator and tests Nos. 3 and 4 by a second operator BY METHOD A Potency found in- I A > Sample test No. 1, test No. 2, test No. 3, test No. 4, No. units per ml units per ml units per ml units per ml - - 1 1540 1560 2 1440 1370 3 1830 1790 - 4 1680 1700 5 1890 1900 6 1720 1760 7 1350 1330 -i - 8 1690 1730 9 1560 1570 10 780 790 11 1990 2000 12 1600 1660 13 1630 1590 14 1270 1280 1320 1300 15 1270 1270 1260 1250 16 1570 1550 1610 1600 17 1640 1660 1550 1510 18 1780 1780 1820 1790 - - - - - - - - - - - - - - - - - - - - - As a further check on accuracy, samples of fermented medium were assayed by both method A and a cylinder - plate method with Bacillus subtilis; for the latter procedure six plate assays incorporating eighteen cylinders of sample and eighteen of penicillin standards (2, 4 and 8 units per m1) were used.The standard error of a single test was approximately 4 per cent. Tests on twenty-four samples gave a mean potency of 1827 units per ml by the B. subtilis method and 1841 units per ml by the iodimetric assay. The closeness of these figures can be regarded as satisfactory evidence of the accuracy of method A. DEVELOPMENT OF DIRECT ASSAY Further work on simplification of the iodimetric method, with the particular object of avoiding the extraction procedure, proved abortive, owing to interference from the relatively large amount of soluble organic matter present in the samples, which led to excessively high blanks.A new approach to the problem was therefore made, in which elimination of inter- ference from iodine-absorbing impurities was attempted by first treating the sample with iodine solution, thereby iodinating the impurities, but leaving the penicillin unchanged ; the penicillin was subsequently determined by a hydrolytic iodination procedure. At the same time, efforts were made to devise a method of adequate accuracy, incorporating all the most desirable features needed in a method to be used for plant control and for assaying experimental laboratory fermentations and flask cultures.These features, which were successfully incorporated into the method, were- (i) Use of a minimum amount of sample, so as to permit (a) small-scale work to be evaluated and (b) a filtered portion of sample to be rapidly obtained. (ii) Avoidance of a solvent-extraction procedure, as this is time-consuming. (iii) Simplicity of procedure and instruments. (iv) Spe~ificity.1~ ,l6452 BETHEL AND BOND : IODIMETRIC METHODS [Vol. 86 (v) Stability of reagents. (vi) Applicability, without modification, to fermentation samples having very varied The more important factors involved in the method are discussed below. compositions. EFFECTS OF CONSTITUENTS OF FERMENTATION MEDIUM ON DIRECT ASSAY- The effects of impurities on iodimetric assay involving inactivation by alkali have been studied by JVildll; he used crude samples of penicillin salts having low potencies and concluded that the method was not entirely satisfactory, owing to absorption of iodine by the impurities, the difficulty in accurately compensating for this absorption in the blank and thc: change in absorption of iodine by the impurities after alkaline hydrolysis, which again could not be compensated for in the blank.The same problem presents itself with fermented media, but in greater degree, as impurities are present in larger proportion than in crude penicillin salts. Tests confirmed the impracticability of applying an iodimetric assay involving inactivation by alkali directly to fermented media. It was apparent, however, that the errors arising from inactivation might be eliminated or decreased by using penicillinase instead of alkali; because of its specific action, penicillinase should inactivate penicillin without having any effect on the other constituents.However, even though the absorption of iodine by these constituents and by the penicillinase can be determined by blank tests, the values are high in relation to the absorption of iodine by the penicillin after inactivation ; further, apart from the undesirability of high blanks, we have been unable to obtain accurate results by this procedure. COMBINATION OF TEST AND BLANKS- In order to avoid difficulties from the impracticability of subjecting both test and blank to the same treatment, we tried a radical departure from the usual iodimetric procedure by combining the blank and test.This was possible because iodine was absorbed nod by peni- cillin, but only by the products of its inactivation. For this purpose, the sample was treated with iodine solution at pH 7.0 to iodinate the impurities as far as was practicable. The absorption of iodine increases with increase in pM, but a pH of 7.0 cannot be exceeded without introducing error from inactivation of the penicillin. The excess of iodine was destroyed by titration with sodium thiosulphate solution, and penicillinass was then added to inactivate the penicillin. Acidification of the solution before iodination of the products of inactivation rendered further absorption of iodine by the impurities negligible. This is essential, as the amount of iodine absorbed at this stage provides the measure of the penicillin in the sample.DEGREE OF ACIDIFICATION- In preliminary experiments it was found for some samples, particularly those of low potency taken early in the fermentation, that, if the test solution was rendered too acid after the preliminary iodination, some iodine w a released; this led to low results. The pH in these tests was about 1-3, which is produced when about 6-5 ml of 1 N hydrochloric acid are added. A pH value of about 2 is produced .with most samples when 5 ml of 0.2 N hydro- chloric acid are added to the test solution. Tests were therefore carried out to determine the effect of pH; thirty-four samples were analysed by the direct procedure at pH values of 1.3 and 2.0 and also by method A.The mean potencies found by the direct method were 3360 (at pH 1.3) and 3560 units per ml (at pH 2-0), and that found by method A was 3580 units per ml. These results show that there is good agreement between results by the extraction method and the direct method carried out at pH 2. PENICILLIKASE- Penicillinase can be prepared by the general method described in the British Pharma- copoeia, but other procedures have been descrjbed.17 to 21 However, for assay purposes it is essential to use a highly potent16 preparation, and we have used nutrient-broth filtrates prepared from BaciZZas cewus. These had potencies of at least 120,000 units per ml per hour a t 20" C, and it was unnecessary to concentrate or purify them further. They would, of course, permit the growth of contaminating micro-organisms, which in turn would cause rapid destruction of the enzyme.It was therefore necessary to add some suitable preservative that would not lead to chemical destruction of the enzyme and would also not interfereJuly, 19611 OF ASSAY FOR PENICILLIN I N FERMENTED MEDIUM 463 with the assay of penicillin. Both toluene and chloroform were found to be suitable for this purpose, and chloroform was more convenient to use, as excess of it sinks to the bottoms of the containers. There was no significant loss of potency after the solution had been stored at 5 ° C for 5 weeks.21 IODINE EQUIVALENT OF PENICILLIN- Since the mode of inactivation of penicillin by penicillinase cannot be identified with alkali inactivation, it was necessary to determine the iodine equivalent of penicillin under the proposed standard test conditions.This was carried out for benzylpenicillin and phenoxy- methylpenicillin as before, recryst allised sodium benzylpenicillin (potency 1670 units per mg) and potassium phenoxymethylpenicillin (potency 1530 units per mg) being used. It was found that the iodine equivalent varied with the concentration and type of penicillin used in the test.13 When the amounts of iodine absorbed were plotted against the concentration of penicillin used, the regression lines were only slightly cuved a t high concentrations of penicillin, but considerably curved at lower concentrations, at which the test is not usually applied. The results obtained for the two penicillins are shown in Table 11.TABLE I1 IODINE EQUIVALENTS OF PEXICILLINS AT VARIOUS CONCENTRATIONS Benzylpenicillin Phenoxymeth ylpenicillin r A > 7 h \ Units equivalent Units equivalent Units present to 1 ml of Units present to 1 ml of per test 0.01 N iodine per test 0.01 N iodine 880 1330 1770 2210 4420 6630 8840 11,050 13,270 521 5.56 572 577 595 606 612 627 635 860 1290 1720 2150 4310 6470 8630 10,790 12,900 541 566 567 567 582 592 598 606 611 TIME REQUIRED FOR INACTIVATION BY PENICILLINASE- Various batches of penicillinase were examined to ascertain that the preferred amount of penicillinase, 0.5 ml, would cause complete inactivation of the optimum amount of penicillin for the test in a suitable short time. The results are shown in Table 111, from which it can be seen that inactivation occurs within 5 minutes under the conditions used; however, in order to have an adequate margin of safety we decided to allow 10 minutes for inactivation.TABLE I11 INACTIVATIOX OF PENICILLINS BY PENICILLINASE Penicillinase used h I \ Amount present Period allowed in 0-5 ml, for inactivation, Batch No. units minutes Benzylpenicillin- r 10 1 20 i :: 44/5 130,000 5 D6 60,000 44/6 130,000 Phenoxymethylpe nicillin- f 5 7 D7 80,000 42/6 245,000 D6 60,000 { 2: } Penicillin added, units 12,400 9900 9900 11,400 6600 11,300 Penicillin found, units 11,700 11,800 11,900 11,800 10,200 10,000 9900 10,000 11,700 11,400 11,500 6700 11,300454 BETHEL AND BOND : IODIMETRIC METHODS [Vol. 86 INDICATOR- In the extraction procedure, starch is normally used as indicator, but we found for some experimental fermentations that the end-point was unsatisfactory. In these instances, however, carbon tetrachloride could be used satisfactorily, and we therefore adopted its use in the direct method so as to make the procedure applicable to all samples.FIRST IODINATION PERIOD-- In order to fix a suitable period for the initial iodination of the broth at pH 7.0, tests were made with iodination periods of 5 and 15 minutes; the results are shown in Table IV. As there was no significant difference between the results after iodination foi- 5 and 15 minutes, the shorter time was adopted for the test. TABLE IV 10,900 + 0.9 10,900 + 0.9 11,000 + 1.9 10,900 + 0.9 11,000 + 1.9 11,000 + 1.9 11,000 + 1.9 10,900 + 0.9 ~ 10,900 + 0.9 POTENCIES FOUND AFTER DIFFERENT PERIODS Potency found by direct method after iodination a t pH 7 for- I 1 Sample 5 minutes, 15 minutes, No.units per ml units per ml V67 2900 2920 v44 1440 1420 v43 2120 2140 v44 1990 2060 V58 1950 1950 OF INITIAL IODINATION Potency found by extraction method, units per ml 2960 1300 2270 2070 1970 Acid-inactivated . . .. 10,800 Acid-inactivated . . .. 8450 Unfermented . . . . 10,810July, 19611 OF ASSAY FOR PENICILLIN I N FERMENTED MEDIUM 455 ASSAY OF SAMPLES HAVING KNOWN PENICILLIN CONTENTS- Checks of the final direct procedure were made on samples of unfermented or inactivated media containing known added amounts of penicillin prepared as before. The results of these tests are shown in Table V and are close to the theoretical values.The mean result for benzylpenicillin is low by 0.08 per cent.; that for phenoxymethylpenicillin is high by 0.12 per cent. These differences are sufficiently insignificant as to need no correction in normal experimental or plant-control work. A further check on accuracy was made by assaying sixteen samples of fermented medium by the direct iodimetric procedure and by a cylinder - plate assay with B. subtilis. The mean of the results by the iodimetric procedure was 1.9 per cent. below that of those by the B. subtilis method. APPLICABILITY- Samples having high potency-The direct method is satisfactory for samples having potencies up to 6000 units per ml. For samples of higher potency, it is necessary to use a proportionately smaller amount of sample than 2 ml or to use 2 ml of a suitable dilution; otherwise, the excess of iodine present is insufficient for complete iodination. Samples having low potency-Assays of samples taken early in the course of fermentation are not usually required.However, if such assays are needed, care is necessary if high concentrations of certain unfermented constituents of the medium, e.g., corn-steep liquor and sugars, are present, as a small amount of absorbed iodine is sometimes released atthe acidification stage of the assay. If this occurs, a blank test must be carried out to measure the amount of iodine released, a correction for which must be applied in the normal test. REPRODUCIBILITY- medium over a short period; the standard error of the results was 1.9 per cent. Replicate tests were made by different operators on fifty-two samples of fermented SPEED- A skilled operator can carry out a single test in about 30 minutes from the receipt of the sample; this represents the minimum time in which a single result can be made available.However, since much of the period is “waiting time,” about six samples of filtered broth per hour can be tested simultaneously (this does not allow for the preparation of reagents and reporting of results). The method is also economical in terms of apparatus and reagents. METHOD B : DIRECT DETERMINATIOK OF BENZYL- OR PHENOXYMETHYLPENICILLIN I N FERMENTED MEDIUM WITH USE OF PENICILLINASE REAGENTS- at 20” C. Penicillinase-This must have a potency of not less than 120,000 units per ml per hour Iodine, 0-02 N-AS for method A.Sodium thiosulphate, 0.01 N-AS for method A. Phosphate bufer solution, 0.01 M (PH 7*O)-Prepare from the sodium salts. PROCEDURE- Accurately transfer, by pipette, 2 ml of filtered sample to a 150- or 250-ml conical flask fitted with a glass stopper. Add about 20ml of 0.01 M phosphate buffer solution and approximately 5 ml of 0.02 N iodine, and set aside for 5 minutes. Destroy the excess of iodine by titration with 0.01 N sodium thiosulphate; use 10 ml of distilled carbon tetrachloride as indicator, and take care not to add excess of the thiosulphate solution. Add 0 6 m l of penicillinase, mix gently, and set aside at room temperature for 10 minutes. Acidify with 5 ml of 0.2 N hydrochloric acid, mix gently, and accurately add 20 ml of 0.02 N iodine.Mix thoroughly, but do not shake (i.e., avoid excessive transfer of iodine to the carbon tetra- chloride), set aside in the dark for 5 minutes, and then immediately titrate with 0.01 N sodium thiosulphate. To determine the blank value (the amount of iodine absorbed by the penicillinase used in the test), proceed exactly as described above, but omit the 2-ml portion of sample; this need be done only once for each batch of penicillinase.456 BETHEL AND BOND IODIMETRIC METHODS [Vol. 86 CALCULATION OF RESULTS- The potency of the fermented broth is calculated from the equation- ( S - A - P ) x F x E 2 Potency, units per ml = - in which A is the volume of 0.01 N sodium thiosulphate required for the test solution, F is the factor of the sodium thiosulphate solution, S is the volume of 0.01 N sodium thiosulphate equivalent to 20 ml of the 0.02 N iodine, P is the difference (in millilitres) between the volume of 0.01 N sodium thiosulphate required to neutralise 20 ml of 0.02 N iodine and that required in the blank determination and E is the penicillin equivalent (in units) of 1 ml of 0.01 N iodine.However, as shown in Table 11, the value of E varies with the type of penicillin assayed and also with the concentration of penicillin in the sample, so that simple application of the above equation is impracticable. For general ease of calculation, it is convenient to can be readily determined. Since construct Tables so that the value of - variations in penicillin equivalent often occur between laboratories, individuals may prefer to make their own assessments, but, if the results in Table I1 are used, the penicillin content of a sample is related to iodine absorbed as shown in Table VI.(These values have been used for calculating the results in this paper.) ( S - A - P ) X E 2 TABL:E VI RELATIONSHIP BETWEEN IODINE ABSORBED AND CONCENTRATION OF PENICILLIN I N SAMPLE Volume of 0.01 N iodine A 3 absorbed Benzylpenicillin, Phenoxymethylpenicillin, Concentration of penicillin in sample ml units per ml units per ml 1 270 270 2 550 550 3 840 840 4 1 150 1140 5 1450 1440 6 1770 1730 7 2080 2040 8 2400 2340 9 2710 2640 10 3030 2940 11 3340 3250 12 3660 3560 13 4000 3870 14 4330 4190 15 4660 4500 16 4990 4830 17 5320 5150 18 5650 5460 19 5980 5780 NOTES ON METHOD B- Samples taken in eady stages of fermentation-Large amounts of unfermented matter are present in samples taken in the early stages of fermentation, and a small amount of iodine is sometimes liberated at the acidification stage of the assay; this is shown by the development of a pink colour in the carbon tetrachloride.With such samples, set the solution aside for 5 minutes after acidification, complete the test in the usual way, and carry out a blank determination as described below. By pipette, place 2ml of the sample in a 150- or 250-ml flask fitted with a glass stopper, and add about 20 ml of 0.01 M phosphate buffer solution. Add 0.5 ml of peni- cillinase, set aside for 10 minutes at room temperature, add excess of 0.02 N iodine (approximately lOml), and set aside for 5 minutes. Destroy the excess of iodine by titration with 0.01 N sodium thiosulphate; use 10 ml of carbon tetrachloride as indicator, and take care not to add excess of the sodium thiosulphate solution.Acidify with 5 ml of 0.2 N hydrochloric acid, shake gently, set aside for 5 minutes, and titrate with 0.01 N sodium thiosulphate.July, 19611 OF ASSAY FOR PENICILLIN I N FERMENTED MEDIUM 457 Let the blank titre of 0.01 N sodium thiosulphate be B ml. Add this result to the test titre, so that the final absorption titre becomes (S - A - P + B) ml, and calculate the potency as above. Treatment of a$@aratus-All apparatus that has contained penicillinase must be steam- heated in an autoclave or heated at 150" C after use to destroy any residual enzyme. CONCLUSIONS Benzyl- or phenoxymethylpenicillin can be determined iodimetrically in fermented medium if (a) partial purification is first carried out by extracting the penicillin into butyl acetate and then into phosphate buffer solution and (b) the blank in the iodination procedure is adjusted to pH 3.6.At this pH inactivation of the penicillin is negligible, and the uptake of iodine by the residual impurities is small and equal to the uptake in the test. These penicillins can also be determined directly and more rapidly, without extraction, by preliminary iodination of the sample; the impurities, but not the penicillin, are iodinated. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. REFERENCES Boxer, G. E., and Everett, P. IN., Anal. Chem., 1949, 21, 670. Beloff-Chain, A., and d'Accadia, F. D., Analyst, 1952, 77, 423. PCnau, H., Hagemann, G., and Saias, E., Ann. Pharm. Frang., 1950,8, 100. Ford, J. H., Anal. Chem., 1947, 19, 1004. Green, N. C., and Marck, P. R., Chem. & Ind., 1959, 1210. Birner, J., Anal. Chem., 1959, 31, 271. Scudi, J. V., and Jelinek, V. C., J . Biol. Chem., 1946, 164, 195. Scudi, J. V., Ibid., 1946, 164, 183. Alicino, J. F., I n d . Eng. Chem., Anal. Ed., 1946, 18, 619. Analysts' Sub-committee of the Ministry of Health Conference on the Differential Assay of Wild, A. M., J . SOC. Chem. Ind., 1948, 67, 90. British Pharmacopoeia, 1958. Weiss, P. J., Antibiotics & Chemotherapy, 1959, 9, 660. Mundell, M., Fischbach, H., and Elbe, T. E., J . Amer. Pharm. Ass., Sci. Ed., 1946, 35, 374. Murtagh, J . J., and Levy, G. B., J . Amer. Chem. Soc., 1945, 67, 1042. Royce, A., Bowler, C., and Sykes, G., J . Pharm. Pharmacol., 1952, 4, 904. Duthie, E. S. J . Gen. Illicrobiol., 1947, 1, 370. Lepage, G. A., Morgan, J. F., and Campbell, M. E., J . Biol. Chem., 1946, 166, 465. Morgan, J. F., and Campbell, M. E., Ibid., 1947, 169, 337. Baudet, P., and Hagemann, G., Experientia, 1954, 10, 374. Kougat, M., Pollock, M. R., and Tridgell, E. J., Biochem. J . , 1956, 62, 391. Penicillin, Analyst, 1949, 74, 550. First received September 2nd. 1960 Amended, March 20th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9618600448

出版商:RSC    

年代:1961

数据来源: RSC

摘要:

July, 19611 OF ASSAY FOR PENICILLIN I N FERMENTED MEDIUM 457 The Trace Determination of Phenyl- and Methylmercury Salts in Biological Material BY J. C. GAGE (Imperial Chemical Industries Ltd., Industrial Hygiene Research Laboratories, The Frythe, Welwyn, Herts. ) A method is described for determining phenyl- and methylmercuric ion in organic material. The sample is made strongly acid with hydrochloric acid and extracted with benzene, and the organic mercury is then re-extracted with aqueous sodium sulphide. After oxidation with acid permanganate, mercury is determined in this solution by a titrimetric procedure; inorganic mercury does not interfere. Recoveries of about 90 per cent. have been obtained when 2 and 20 p.p.m. of phenyl- and methylmercury acetates were added to biological material.The method is capable of measuring concen- trations of mercurials well below 1 p.p.m. AN investigation in this laboratory into the metabolic fate of phenyl- and methylmercury salts required a method of determining trace amounts of these substances in animal tissue in the presence of inorganic mercury. Methods are available for examining crops treated458 GAGE: THE TRACE DETERMINATION OF PHENYL- AND [Vol. 86 with organic mercurial fungicides, but these are restricted to the determination of total mercury. More specific methods have been suggested, based on ultra-violet spectrophoto- metryl and polarography,2 but these have not been elaborated for a range of biological materials. Miller, Lillis and Csonka3 have described a method in which the sample is oxidised with alkaline permanganate, after which an arylmercury salt may be extracted with a solution of dithizone in chloroform.This method is nlot very sensitive, as the yellow colour of the complex formed from dithizone and phenylmercuric ion is not intense, and it is necessary to measure the green colour of the unchanged diithizone. Moreover, the method is stated to be unsuitable for alkylmercury salts. A method for determining phenyl- and methylmercury salts in biological n naterial has been developed in which the substance is extracted by benzene from a homogenate of the sample that has been made strongly acid with. hydrochloric acid. The benzene solution is shaken with a dilute aqueous solution of sodium sulphide, with which these organic mercurials react to form water-soluble complexes? The aqueous layer is then oxidised with acid permanganate, and the mercury content is determined by a modification of Fabre, Truhaut and Boudhe’s highly sensitive titrimetric m e t h ~ d .~ REAGENTS- METHOD Sodium sulphide solution, 1 per cent. w/v. Hydroxylamine hydrochloride solution, 20 ;@er cent. w/v, Urea solution, 10 per cent. w/v. EDTA solution-Prepare a 2.5 per cent. w/v solution of the disodium salt of ethylene- Sodium acetate solution, 50 $er cent. zu/zu--Prepare from sodium acetate trihydrate. Dithizone solution-Prepare a stock solution containing 20 mg of dithizone in 100 ml of Prepare the working solution Standard mercury solution-Prepare a stock solution by dissolving 67.7 mg of mercuric For use, dilute 1 ml of the stock solution to diaminet etra-ace tic acid.carbon tetrachloride ; store this solution in a refrigerator. when required by diluting 1 ml of the stock solution to 20 ml with chloroform. chloride in 1 litre of 0.5 N hydrochloric acid. 50 ml with water; this solution contains 1 pg of mercury per ml. EXTRACTION OF ORGANIC MERCURY- Prepare a homogenate containing 5 per cent. of the tissue in water, measure 20 ml into a stoppered tube, and add 4 ml of hydrochloric acid, sp.gr. 1.18, and 20 ml of benzene. Shake the tube for 5 minutes, allow the layers to sepamte, and spin in a centrifuge if necessary to obtain a clear upper layer. Transfer 15 ml of the upper benzene layer to a second stoppered tube, add 10ml of sodium sulphide solution, .and shake for 5 minutes.Allow the layers to separate, spin in a centrifuge if necessary t o obtain a clear lower layer, and transfer the lower layer as completely as possible to a third stoppered tube. Shake the benzene layer with 5ml of water, allow to separate, and add the lower aqueous layer to the contents of the third tube. OXIDATION WITH ACID PERMANGANATE- potassium permanganate and 1 drop of octanol. boiling-water bath for 1 hour. chloride solution to decolorise, and cool to room temperature. To the aqueous sulphide solution add 2 m:l of 50 per cent. v/v sulphuric acid, 0.5 g of Place the loosely stoppered tube in a Allow to cool a little, add 3ml of hydroxylamine hydro- DETERMINATION OF MERCURY- To the oxidised solution add 2 ml of urea solution and 1 ml of EDTA solution, adjust the pH to 1-5 with sodium acetate solution (use an external indicator paper), and add 0-5 ml of chloroform; this is the test solution.Prepare a control solution by mixing 15 ml of water, 2 ml of 50 per cent. v/v sulphuric acid, 0.5 g of potassium permanganate, 3 ml of hydroxyl- amine hydrochloride solution and the remainder of the reagents as above; adjust to the same pH. From the burette, add the diluted dithizone solution in small portions to the test: solution; replace the stopper, and shake the tube vigorously after each addition. If mercury is present, the separating organic layer will have an orange colour. Continue the titration until the chloroform layer has a greyish colour, intermediate between the orange of the mercury complexJuly, 19611 METHYLMERCURY SALTS IN BIOLOGICAL MATERIAL 459 and the green of unchanged dithizone.Record the volume of dithizone solution used, and add this same volume to the control solution. Titrate the control solution with standard mercury solution; add the titrant in small portions, with vigorous shaking after each addition, until the colour of the chloroform layer that separates matches that of the layer in the test solution. The volume (in millilitres) of standard mercury solution used is equal to three- quarters of the amount of organic mercury (in micrograms) present in the original sample. RESULTS AND DISCUSSION OF THE METHOD To determine the recoveries of phenyl- and methylmercury salts by the proposed method, known amounts of phenyl- and methylmercury acetates were added to homogenates of normal rat tissue or to rat urine.Solutions containing 1 mg of the acetates per ml were prepared in acetone, and these were diluted with water just before addition to the samples. The results in Table I show that recoveries were low by up to 15 per cent. and that this deficit does not seem to be affected by the concentration of the mercury compound in the tissue; the recovery from liver was the same at concentrations of 2 and 20 p.p.m. The method is sufficiently accurate for metabolic studies on organic mercurials and is capable of measuring concen- trations well below 1 p.p.m. TABLE I RECOVERY OF PHENYL- AND METHYLMERCURY ACETATES ADDED TO RAT TISSUE AND URINE Methylmercury acetate A r Amount of Amount Amount Sample sample taken added, found, Recovery, la Pg % Liver 0.5 g 10 9.2 92 Liver 1.0 g 2 1.8, 1.9 91 Brain 1.0 g 2 1.8, 1.7 87 Kidney 1.0 g 2 1.8, 1.8 90 Urine 10 ml 10 8.8 88 Phenylmercury acetate h r 1 Amount Amount added, found, Recovery, Pg PLg % 10 8.8 88 2 1.75, 1-75 88 2 1.75, 1-65 85 2 1.85, 1-75 90 10 9.5 95 Inorganic mercury present in the tissues is not extracted into benzene under the conditions of the method and does not interfere.It has not usually been found practicable to measure the amount of inorganic mercury in the sample by analysis of the aqueous layer from the first extraction with benzene, as this contains a proportion of the organic mercury dissolved in the benzene, which remains emulsified and is not easy to separate. If the ratio of inorganic to organic mercury in the sample is large, the error from the residual organic mercury in this layer may not be significant, or it may be possible to remove the interference by washing the aqueous layer with benzene. In general, however, it is more convenient to determine the total-mercury content of a second sample and to obtain the inorganic-mercury content by difference. Although the proposed method has been developed for determining phenyl- and methyl- mercury salts in animal tissues, there seems to be no reason why it should not be applicable to the determination of alkyl- and arylmercury compounds in vegetable samples. Technical assistance in this investigation was provided by Mr. P. Salt and Mr. W. B. C. Lyle. REFERENCES 1. 2. 3. 4. 5. Eldridge, A., and Sweet, T. R., Anal. Chem., 1956, 28, 1268. Kotakemori, M., and Handa, H., Ann. Rep. Takamine Lab., 1956, 8, 231; Anal. Abstr., 1958, Miller, V. L., Lillis, D., and Csonka, E., Anal. Chem., 1958, 30, 1705. SDorek, K. F., Analyst, 1956, 81, 478. Fabre, R., Truhaut, R., and Boudkne, C., Ann. Biol. Clin., Paris, 1958, 16, 286. 5, 4300. Received March 2nd. 1961

ISSN:0003-2654    

DOI:10.1039/AN9618600457

出版商:RSC    

年代:1961

数据来源: RSC

摘要:

460 MARZYS : THE SPECTROPHOTOMETIUC DETERMINATION OF SULPHA [Vol. 86 The Spectrophotometric Determination of Sulphadiazine, Sulphamerazine and Sulphathiazole in Mixtures BY A. E. 0. MARZYS (World Health Organisation, Central Pirblic Health Laboratory, Tehran, I r a n ) A rapid method is described for the spectrophotometric determination, without prior separation, of sulphadiazine, sulphamerazine and sulphathiazole present in one solution. Sulphadiazine is determined by the Association of Official Agricultural Chemists’ method, which involves use of 2-thio- barbituric acid and measurement a t 532 mp of the colour developed. Direct optical-density measurements on a solution of the mixture in hydrochloric acid are made at 243 and 280 mp, and the sulphamerazine and sulphathiazole contents are then calculated by solving simultaneous equations in which the measured values of EF& and the sulphadiazine content found have been substituted.The method has been applied to a range of prepared and commercial mixtures. THE use of sulphonamide mixtures has been fairly common since sulphonamide therapy was first introduced, and the problem of determining individual components in such mixtures has been tackled in several ways. One frequently used approach consists in paper-chromato- graphic separation and subsequent titrhetric or colorimetric determination of the separated sulphonamides. Two official methods1s2 are based on this type of procedure; the technique is naturally somewhat lengthy, and results are not accurate unless averages of multiple determinations are taken.One common mixture of sulphonamides contains sulphadiazine, sulphamerazine and sulphathiazole, and numerous preparations incorporating these three compounds are available in various countries. The British Pharmaceutical Codex3 lists a formula for trisulphonamide tablets consisting of these three components, but the assay method described is for total- sulphonamide content only. The object of this work was to find a rapid and reasonably accurate method for determining the individual components of preparations containing these three sulphonamides without prior separation. In the past, single sulphonamides have been determined by making use of their pro- nounced absorption maxima in the ultra-violet r e g i ~ n . ~ Banes5 took advantage of the particularly favourable difference between the absorption spectra of sulphadiazine and sulphathiazole in acid ; he devised a simultaneous spectrophotometric method based on measuring the optical densities of the solutions at 239 and 280mpJ at which the two com- ponents have their absorption maxima, and solving simultaneous equations in the usual manner.Application of a similar procedure to a system also containing sulphamerazine was prevented by the fact that sulphadiazine and sulphamerazine have closely similar spectra. To overcome this difficulty, Thomas and Lagrange6 proposed a method based on the colours produced when the aminopyridine groups of sulphamerazine and sulphadiazine were allowed to react with resorcinol and sulphuric acid; these colours had absorption maxima at different wavelengths.Sulphathiazole was determined by direct optical-density measure- ments at 280mp in hydrochloric acid, and corrections were made for sulphadiazine and sulphamerazine. However, when this method was tried, Thomas and Lagrange’s claims as to stability of solutions and lack of interference by sulphathiazole could not be substantiated, and the precision and accuracy were poor; further, two different solutions of the sample were needed, which fact did not make the method rapid. Attention was then turned to the AssociaiLion of Official Agricultural Chemists’ method’ for determining sulphadiazine in the presence #of other sulphonamides by means of its colour reaction with 2-thiobarbituric acid. In mixtures of sulphadiazine and sulphamerazine, the latter is determined by measuring the total absorption in 0.1 N hydrochloric acid and then subtracting the absorption due to sulphadiazine (known from its colorimetric determination).It seemed reasonable to suppose that, by a combination of this method with that described by BanesJ5 the three sulphonamides could be determined in the same solution. EXPERIMENTAL All optical-density measurements were made in matched 1-cm fused-silica cells with a The Beckman DU spectrophotometer, and a cell correction was applied when necessary.July, 19611 DIAZINE, SULPHAMERAZINE AND SULPHATHIAZOLE IN MIXTURES 461 ultra-violet spectra were plotted for solutions of the sulphonamides in 1 and 0.1 N hydrochloric acid, and little difference was found between the results at these two concentrations of acid.For convenience of combination with the thiobarbituric acid method for sulphadiazine, 0-1 N hydrochloric acid was used. The absorption spectra for the three sulphonamides at a concentration of 10 pg per ml (1 mg per 100 ml) are shown in Fig. 1. Maximum absorp- tion was at 243 mp for sulphadiazine and sulphamerazine and a t about 280 mp for sulpha- thiazole. With the instrument used, the values found for E:2m were- Sulphadiazine Sulphamerazine Sulphathiazole E:tm at 243 mp . . * . 568 579 119 Et'& a t 280 mp . . .. 119 128 492 Additivity of optical densities was ascertained, and solutions of all three sulphonamides were found to obey the Beer - Lambert law at the wavelengths indicated above; for concen- trations of sulphonamide between 0 and 25 pg per ml (0 to 2.5 mg per 100 ml), graphs of concentration against optical density were linear and passed through the origin.0.6 0.4 x c W U cd Y ro .- - .- U 0" 0.2 a 20 240 260 280 300 320 Wavelength, my Fig. 1. Ultra-violet absorption spectra of solutions containing 1 mg of sulphonamide per 100 ml in 0.1 N hydrochloric acid: curve A, sulphathiazole; curve B, sulphamerazine ; curve C , sulphadiazine Tests of the A.O.A.C. thiobarbituric acid method for sulphadiazine confirmed all the findings of a previous collaborative study.* With the reagents and apparatus used, maximum absorption of the colour developed was at 533 instead of 532 mp, and recrystallisation of the 2-thiobarbituric acid was unnecessary, Instead of glass-stoppered tubes, which have to be weighed before and after heating, it was found more convenient to use 10-ml calibrated flasks. A l-ml portion of the sample or standard solution was placed in the flask, the contents of which were then diluted to the mark with buffered thiobarbituric acid solution.The flask was placed in a bath of boiling water, and, when the solution had attained the temperature of the bath, the stopper was inserted; the solution was then heated in the bath for 1 hour. When the flask was cool, any moisture lost was replaced by diluting to the mark with water. The amount of reagent present in relation to the amount of sample solution was therefore slightly less than in the A.O.A.C. method, but this did not affect reliability in any way. The influence of extraneous factors on development of colour and extinction coefficient was over- come by including standards with every batch of sample solutions, and results were calculated as in the A.O.A.C.m e t h ~ d . ~ No interference from sulphamerazine or sulphathiazole was noticed when these compounds were present in proportions likely to be encountered in commercial mixtures.462 MARZYS 1 THE SPECTROPHOTOMETR[C DETERMINATION OF SULPHA- [VOl. 86 If the optical densities of a solution containing sulphadiazine, su1phameri:izine and sulphathiazole in 0.1 N hydrochloric acid at 243 and 280 mp are Sz43 and s280, respectively, then, since the optical density of the mixture is equal to the sum of those of its constituents- 1000S,4, = 568d $- 579m + 119t 1000S280 5 119d 4- 128m + 4925 where d, m and t are the concentrations of sulphadiazine, sulphamerazine and sulphathiazole, respectively, in milligrams per 100 ml.The value of d is known from the determination of sulphadiazine, so that, if (lOOOS,,, - 568d) and (1000S2,0 - 119d) are denoted by S’243 and S’280, respectively, the above equations can be written as- S’243 = 579m + 11% S’280 = 128m + 4925 from which the values of m and t can be calculated. 1 19st280 - 492S’2, (128 x 119) - (579 x 492) m= t = 579S’,,o - 128St,,, (579 x 492) - (128 x 119) The values of the extinction coefficients used in these calculations apply to one particular instrument and time and must be periodically Checked, although differences should be small. METHOD APPARATUS- Spectrophotometer-Use an instrument provided with a tungsten lamp for measurements in the visible region, a hydrogen-discharge tube for measurements in the ultra-violet region and fused-silica or quartz 1-cm cells matched in pairs.REAGENTS- Citrate bufer and 2-thiobarbituric acid solutions-Prepare as described‘ in A.O.A.C. method No. 32.285; recrystallisation of laborattory-reagent grade 2-thiobarbjturic acid is not necessary if the blank is treated in exactly the same way as the standard and sample solutions. described7 for sample solution A in A.O.A.C. method No. 32.286; use 0*1000 g each of pure sulphadiazine, sulphamerazine and sulpha- thiazole. For the determination of sulphamerazine and sulphathiazole, place 5 n-tl of each solution in separate 25-ml calibrated flasks, and dilute each to the mark with 0.1 N hydro- chloric acid.Each of these solutions contains 10 pg of sulphonamide per 11-11 (1 mg per 100 ml). For the determination of sulphadiazine, place 5 ml of the sulphadiazinle solution in a 10-ml calibrated flask, and dilute to the mark with 0.1 N hydrochloric acid. This solution contains 25 pg of sulphadiazine per ml (2.5 mg per 100 ml). Standard sulphonamide solutions-Prepare PROCEDURE FOR SULPHADIAZINE- Determine sulphadiazine as described in A.O.A.C. method No. 32.286,7 but use 10-ml calibrated flasks instead of tubes fitted with ground-glass stoppers. Dilute 1-ml portions of sample and standard solutions and 0.1 N hydi-ochloric acid (blank) to 10 ml with buffered 2-thiobarbituric acid solution. Place the flasks in a bath of boiling water, insert the stoppers when the solutions have attained the temperature of the bath, and then heat for 1 hour.Remove the stoppers to prevent them from sticking, allow to cool, and add water, if necessary, to compensate for any loss of moisture. PROCEDURE FOR SULPHAMERAZINE AND SULPHATHIAZOLE- By pipette, place 5 ml of sample solution A (prepared as described7 in A.O.A.C. method No. 32.286) in a 25-ml calibrated flask, dilute to the mark with 0.1 N hydrochloric acid, and measure the optical density of the solution at 1243 and 280 mp against 0.1 N hydrochloric acid. Determine the values of Eitm for the three sulphonamides from measurements made on the standard solutions containing 1 mg of sulphonamide per 100 ml. Substitute the values so obtained in the above equations, and calculate the concentrations of sulpha- merazine and sulphathiazole in the sample solution.Hence deduce the content: of each sulphonamide in the original sample. Continue as in the A.O.A.C. method.July, 19611 DIAZINE, SULPHAMERAZINE AND SULPHATHIAZOLE IN MIXTURES 463 RESULTS AND DISCUSSION OF THE METHOD The proposed procedure was tested on four mixtures containing known amounts of the three sulphonamides; the results are shown in Table I. A variety of commercial products, both locally prepared and imported from various countries, including tablets as well as liquid preparations (suspensions) was then analysed, and some of the results are shown in Table 11. The accuracy of the results is as good as can be expected from a simultaneous spectro- photometric procedure for mixtures.For commercial preparations, tolerances in the amounts present must naturally be added to the analytical error. Some interference occurs in the colorimetric determination of sulphadiazine when the other two sulphonamides are present in large excess, e.g., 20 times the amount of sulphadiazine (see mixture No. 4 in Table I). This, however, is a proportion unlikely to be found in commercial preparations. TABLE I SULPHONAMIDE CONTENTS FOUND IN PREPARED MIXTURES Sulphadiazine content- Sulphamerazine content- Sulphathiazole content- -------7 7- 7 - 7 Mixture present, found, present, found, present, found, NO. mg mg mg mg mg mg 1 175 175 175 170 175 176 2 250 248 150 154 100 98 3 125 126 50 48 325 323 4 25 28 300 304 175 168 TABLE I1 SULPHONAMIDE CONTENTS FOUND IN COMMERCIAL TRISULPHONAMIDE PREPARATIONS The figures for samples A, C and D are expressed as milligrams of sulphonamide per tablet and those for sample B as milligrams of sulphonamide per fluid drachm (3.6 ml) Sulphadiazine content- Sulphamerazine content- Sulphathiazole content- r 7- r “g mg mg mg mg mg A (tablets) .. 185 183 130 135 185 196 B (suspension) 185 184 130 127 185 186 C (tablets) . . 100 96 200 198 150 157 D (tablets) . . 185 75 130 Nil 185 30 1 Sample stated, found, stated, found, stated, found, The four commercial products tested in Table I1 were from three different countries. Sample D was obviously a falsification, in which the amount of sulphathiazole (the cheapest of the three sulphonamides) was double that indicated, the more expensive sulphadiazine was present in less than half the amount stated and the most expensive sulphamerazine was absent.The proposed method is rapid and simple ; it fulfils the requirements of a routine procedure for determining the three individual sulphonamides in their most common combination. Semi-skilled operators have analysed, without difficulty, a batch of four samples in duplicate during one morning. I thank Dr. A. H. Taba, Director, World Health Organisation, Regional Office for the Eastern Mediterranean, and Dr. M. Zoriassatein, Director, Central Public Health Laboratory, Tehran, for permission to publish this paper. I also thank Dr. K. Yazdi and the staff of the Pharmaceutical and Toxicological Section of the Laboratory for their assistance in experi- ment a1 work. REFERENCES Commonly used excipients have caused no interference. 1. 2. 3. 4, 5. 6. 7. 8. “The Pharmacopoeia of the United States of America,” Fifteenth Revision, Mack Publishing Co., “The National Formulary,” Tenth Edition, The American Pharmaceutical Association, Washing- “British Pharmaceutical Codex, 1959,” The Pharmaceutical Press, London, 1959, p. 1150. Ciminera, J. L., and WiIcox, P. W., J . Amer. Pharm. Ass., 1944, 33, 85. Banes, D., J . Ass. Off. Agric. Chevn., 1948, 31, 653. Thomas, J., and Lagrange, G., J . Pharm. Belg., 1951, 6 , 355. Horwitz, W., Editor, “Official Methods of Analysis,” Ninth Edition, The Association of Official Agricultural Chemists, Washington, D.C., 1960, p. 526. Conroy, H. W., J . Ass. Off. Agric. Cltem., 1954, 37, 697. Easton, Pennsylvania, 1955, p. 758. ton, D.C., 1955, p. 723. Received January 19th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9618600460

出版商:RSC    

年代:1961

数据来源: RSC