ISSN:0003-2654    

DOI:10.1039/AN95479FX043

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479BX045

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479FP117

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

... THE ANALYST XlllETALLURGICAL CHEMIST. Competent youngManalyst required with knowledge of ferrous and non-ferrous analysis and the use of modern techniques. Excellentopportunity for carrying out interesting work connected witha variety of research and development projects. The positionis permanent and progressive. Some knowledge of platingan advantage. Should have completed National Service.Superannuation scheme in operation. Send full details,including salary required to Box No. 3868, THE ANALYST,47, Gresham Street, London, E.C.2.NALYST required to take charge of Department dealingpesticide residues on crops, atmospheric contamina-tion and mion analysis of organic compounds. Salary up toL1,OOO per annum according to qualifications and experience.Write (quoting No.933), stating age, fullest particulars andsalary required, to Personnel Manager, Pest Control Limited,Bourn, Cambridge.DERBYSHIRE COUNTY COUNCILDEPUTY COUNTY ANALYSTAPPLICATIONS for the above appointment are invitedfrom persons who are qualified in accordance with thePublic Analysts Regulations.J.N.C. conditions of service; salary Grade C (L1,050 x 50-E1,250) ; post pensionable; medical examination.Apply, giving. age, appointments held (with dates andsalaries) and listing three referees, to The County Analyst,St. Mary’s Gate, Derby, by 30th September.Canvassing disqualifies.UBLIC ASALYST’S DEPARTMENT CITY OFPPORTSMOUTH, invites applications fo; the post ofAssistant Analyst from candidates with B.Sc., and/orA.R.I.C., qualification(s).Salary L580 rising to L8lO byannual increments. The candidate’s qualifications andexperience will be taken into account in determining theinitial salary to be offered. Satisfactory medical examinationrequired. Form of application and further details may beobtained from the Public Analyst’s Department, TrafalgarPlace, Clive Road, Portsmouth.ASSISTANT ANALYST (age 19-27) required for BritishWelding Research Association. O.N.C. level or equivalent.Several years’ experience of metallurgical analysis required.Must be careful and accurate. Good prospects. ApplySecretary, B.W.R.A., 29, Park Crescent, London, W.l.HER MAJESTY’S COLONIAL SERVICEVACASCY exists for a Chemist, Medical Department,AHong Kong. Age limits 35 years.Honours Degree of British University inChemistry and/or Associateship or Fellowship ot R.I.C.;three years post graduate experience.Assist Government Chemist analytical and con-sulting work concerned with public health; control ofnarcotic drugs, strategic materials, dangerous cargoes;commercial analytical work, toxicology etc.Terms of Appointment. Pensionable with salary in thescale L750-L1,406 p.a. plus pensionable overseas allowanceof between L2lO-L280 p.a. A temporary non-pensionablecost of living allowance is also payable. Free passages onceeach way each tour; leave at rate of 1 day for every 7 daysof resident service; quarters, if available, at rental of notmore than & of basic salary; income tax at local rates.Apply in writing to the Director of Recruitment, ColonialOffice, Great Smith Street, London, S.W.l., giving brieflyage, qualifications and experience.Mention the referencenumber (BCD.117/51/018).Qualifications.Duties.~~ ~WOOLWICH POLYTECHKIC, LOKDOK, S.E.18Principal:-J. S. Tait, Ph.D., BSc., A.K.T.C., M.I.E.E.,A.M.1.Mech.E.PPLICATIONS invited for a Senior Lecturer in OrganicAChemistry from candidates with high academic andresearch qualifications, and able to teach to Special (Honours)Degree standard.Salary scale L1,065-L25-L1,215, plus London allowance.Particulars and application form from Clerk to Governors,to be returned by 28th September, 1954.RADUATE BIOCHEMIST or CHEMIST with bio-Gchemical experience required for Medical ResearchLaboratory of Pest Control Limited, Cambridge (Member ofFison Group).Duties mainly on enzyme (cholinesterase)and toxicological studies. Sound knowledge of manometrictechniques is essential. Salary according to age and ex-perience. Write (quoting No. 946), stating age, qualifications,previous appointments and salary required, to PersonnelManager, Pest Control Ltd., Bourn, Cambridge.ARGE FUEL TESTING LABORATORY on North- L East Coast requires Graduate as assistant to Chief Chemist.Applications are invited from holders of B.Sc. or M.R.I.C.,preferably aged 25-30. Write to Box No. 3869, THE ANALYST,47, Gresham Street, London, E.C.2.ICROANALYST. May & Baker Limited, Dagenham,MEssex, have a vacancy for an analyst of B.Sc. or A.R.I.C.standard in the Microanalytical Department of their ChemicalResearch Organisation.The duties will include not onlysupervision of quantitative microanalysis of organic coni-pounds by established procedures, but also experimentalwork designed to introduce new and improved methods andto widen the scope of the laboratory. Applicants must havea good knowledge of organic chemistry and experience inmicroanalysis or at least a real aptitude and liking for micro-scale techniques, so as to take advantage of the opportunityfor acquiring skill in this specialised branch of analysis whichis vital to research. Apply initially in writing to the PersonnelOfficer.KALYST. Applications are invited for a senior super-Avisory post in a modern analytical laboratory (chieflyinorganic) in the inner London area.The laboratory isattached to a large manufacturing organisation and the postis permanent and pensionable. Our own staff are aware ofthe vacancy. Please send brief details of age and experience,with some indication of salary expected, to Box No. 796,c/o J. G. King & Son, 150, Fleet Street, London, E.C.4.HIEF CHEMIST. Applications are invited from candi-Cdates, having the undermentioned qualifications, for thepost of Chief Chemist at a mine in Northern Rhodesia.BSc. degree, or equivalent, with chemistry as a majorsubject is essential, preferably with honours standard. Goodall-round experience in analytical practice, especially in theuse of up-to-date techniques and instruments is required.Particular experience in metallurgical analysis is desirable,and experience in analytical work connected with base metalmilling, smelting and/or leaching will be an added qualifica-tion.Applicants should have had experience in administra-tion of a laboratory and control and direction of staff.Basic salary kl,200 to L1,400; plus cost of living allowanceand bonus, at present k120 per annum and 15% of basicsalary respectively. Contributory Pension Scheme. FreeLife Assurance, unfurnished housing and medical services.Passages of successful applicant, with wife and dependentchildren paid from present home to Northern Rhodesia.Annual leave on full pay, 55 days, which may be accumulatedfor three years. In addition, 5 days local leave per annuni.Candidates should submit full particulars of career, age andmarital status, enclose a recent photograph, and state whenthey would be available to take up the appointment.Applications should be addressed by airmail to :-TheSecretary, The Rhodesia Broken Hill Development CompanyLimited, P.O.Box 172, Kitwe, Northern Rhodesia.NALYTICAL CHEMIST required for large modernAresearch establishment in Northern Rhodesia engaged inresearch in the base metal industry. Applicants shouldpossess a University Degree or its equivalent in Chemistry,and a minimum of 3 years’ practical analytical experience.Particular experience in the field of spectrographic analysisis required. Accommodation available for single men.Applications from married men can only be considered if theyare prepared to live in Single Quarters for approximately8 months.Starting salary $38 per month plus variablecopper bonus (at present approximately 60% of basic salary)and cost of living allowance (at present L l O 8s. Od. per month).Contributory Pension Scheme and generous leave conditions.Applications stating age, qualifications, experience, avail-ability should be addressed to Box No. 950, Dorland Adver-tising Ltd., 18/20, Regent Street, London, S.W.l.HEFFER’S BOOKSHOPScientific andTechnical BooksNew and SecondhandPetty Cury, Cambridgxvi THE ANALYSTRepresentatives for sales and service:Mancherter area:-Messrs. A. M. Lock & Co., Ltd.. Crompton Street, Chadderton,Oldham, Lancs. North of England:-Mersrr.Electricalr, Ltd., 14 Claremont Place,Newcastle upon Tyne. 2. Scotland:-Mersrs. Atkins, Robertson & Whiteford,AnexaminationbyOPACIMETERSSPECTROPHOTOMETERSLIGHT OPERATED RELAYSREFLECTOMETERS. etc., etc.M. Filhol“Vegetable Golouring Matters:-M . Filhol has been engaged in theexamination of vegetable colouringmatters, and has discovered somefacts which he now publishes asbriefly as possible, intending to giveall the details in a longer memoir.There exists in nearly all flowers,says M . Filhol, a substance which isscarcely ccloured when in solution inacid liquids, but which becomes of abeautiful yellow colour when actedupon by alkalis.” (Chemical News,1860 (April 14), I, 228).LABORATORYM. FILHOL’S DISCOVERIES were made nearlya hundred years ago.Natural dyes have long been superseded asindicators by the highly purified synthetic dyecompounds used to-day for colorimetric measure-ments of hydrogen ion concentration, oxidation-reduction balance and adsorption capacity.Theoryand practice in these fields are explained in threeB.D.H. booklets - ‘ pH Values ’, ‘ The Colori-metric Determination of Oxidation-ReductionBalance ’ and ‘Adsorption Indicators ’-which maybe obtained free on request.BODOH C H E M I C A L STHE BRITISH DRUG HOUSES LTD. B.D.H. LABORATORY CHEMICALS GROUP ’ P O O L E . DORSETICCIPIITHEPORTABLE CO LORI I----This unique instrument provides a simple photo-electric means of accurately assessing the colour densityof a liquid.It combines in a comparatively smalrobust case the Colorimeter measuring block, micro-ammeter and power unit and i s therefore quite com-plete. It employs standard test tubes, can beoperated twenty-four hours per day without over-heating and is unaffected by external lighting.A wide range of colour filters is available for an . infinite number of determinations. An outstandingphoto-electric device for application in many spheres,both for research and routine tests.The “EEL” Portable Colorimeter may be used withthe minimum of tuition o r skill but i t s simplicity ofconstruction and operation do not affect the accuracyof results, rather the reverse, but do enable i t s supplyat a low price.Send for full details from*~ -_FLAME PHOTOMETERSPHOTOMETERSDENSITOMETERSCOLORIMETERSEVANS ELECTROSELENIUM LTTHE ANALYST x s iANALYTICALCHEMISTRY .....ISpublished' monthly. .* * by theAMERICANCHEMICALSOCIETYas a service toYOURPROFESSIONproviding annuallyabout1800 PAGEScontaining over500TECHNICALPAPERSof importance toyour analytical ... workat the .-.subsiription pricesand.. 'postage rates of:I year ... $5.202 years ... $9.403 years ... $13.60. *....American Chemical Society1155 Sixteenth St., N.W.Washington 6, D.C., U.S.A.W. Heffer & Sons, Ltd.Cambridge, England___-_or _____LATEST ADDITIONS TO 1954 LIST 1I1 -Amino- 4-paminophenol4-Amino-4-chlorodiphenylnaphthalene . . . . 7 0 - DHCl . . . . . . 4 0 - DBenzalazine ( 0 7 ~ ' ~ ~ ) . . . . 4 - DBenzene sulphonic acid ( o S 0 , , ) 25 - H4-Benzoyl amino-2,5-di-methoxyaniline diazotate 6 - DBis- cyclohexanone-oxalyldihydrazone . . 13 - DChloranilic acid . . . . 2 0 - DDecane (petroleum fraction3,3-Dianisole bis 4,4- (3,5-diphenyl) tetrazoliumI ( I ~ 1s-j C.) . . . . 2 0 - Kchloride . . .. 3 6 - G2,2'- Dihydroxy- 6,6'-5,6-Dimethyl-1 ,lo-phen-N, N '-Dimethyl-phenyl3,4-Dinitro-benzoic acid . . 12/- GGallacetophenone (2,3,4-Trihydroxyacetophenone) 161- Ddinaphthyl disulphide 22 - Ganthroline . . .. 81- Dhydrazine . . . . 30/- D. . 251- DMercury diethyl . . _ _ 15/- DMercury diphenyl . . . . 251- DGlyceric acid . . . .Methyl pyruvate . . .. 481- DPalmitic acid (c)8(',, - ) . . 6/- Hiso-Pentane (boiling range1,3,5-Triacetoxy-benzene . . 671- D2,4,6-Tribromo-diphenyl . . 91- GCelite Filter Aids available in1-lb., 5-lb. and 50-lb. packs.EDTA-metal complexes avail-able in experimental quant-ities.Ion exchange resins andPrecious Metal Catalystsalso available.Write now for our 1954 Cata-logue and recent Supple-ment to :-27-31- 3.) . . . . . . 51- H1. LIGHT & CO LTD 1POYLE, COLNBROOK 1SLOUGH

ISSN:0003-2654    

DOI:10.1039/AN95479BP127

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

SEPTEMBER, 1954 Vol. 79, No. 942 THE ANALYST Editorial THE PRESENTATION OF PAPERS AT MEETINGS IT is now some time since a suggestion was made by the Committee of the Microchemistry Group that many authors of papers, on receiving an invitation to present their work at a meeting, would welcome some guidance on the best method of doing so, and that it was desirable for authoritative written advice to be at hand for their information and instruction. To this end, the Group Committee, on behalf of the Publication Committee, drafted some notes and instructions. This first draft received very full consideration by the Publication Committee, during which many of the difficulties and dangers of issuing precise instructions for the presentation of every variety of paper by speakers of differing ability and experience were discussed.In addition to papers already written, submitted for publication and selected for presentation by the Publication Committee, there was the important class, as yet uncommitted to paper and dealing with work still in progress, that was brought before a meeting by special invitation, more with the idea of stimulating interest in new work and inviting discussion than with recording a finished piece of work. The differing degrees of ability and experience of the speakers themselves had also to be taken into account. Nevertheless, despite these irreconcilable divergencies of subjects and speakers, the manner in which papers should best be presented to an audience was held to be of sufficient importance to warrant an exposition of its general principles in a form useful to all speakers, appropriate to any subject and suitable for adoption at all meetings throughout the Society. A statement of these principles has now been prepared and is printed in this issue. During the discussions that took place in the preparation of this disquisition, one member of the Publication Committee, himself a practised and well-known speaker at meetings of the Society, put forward suggestions for fuller treatment of certain aspects of the subject, on which he held decided opinions. These he was invited to record in the form of an essay, which also appears in this issue. Both of these writings are of importance to intending lecturers; in one they will be instructed in the minimum requirements of the Society and in the other they will be able to learn something of the successful speaker’s art. 529

ISSN:0003-2654    

DOI:10.1039/AN9547900529

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

530 COMMUKICATIONS AT MEETINGS IS GENERAL [VOI. 79 Communications at Meetings in General COMMUSICATIOSS to meetings of scientific and other learned societies may range from the telegraphic transmission in a few ininutcs of a worker’s own experimental finhngs to the meticulously worded formal discourse on some solcrnn and ccrcmonial occasion. Intermediate between these extremes lip: the longer presentation of a series of research achievements, often those of a team rather than an individual, and the survey or review type of lecture in which tlie author has, qzin expert, been invited to combine historical record with critical assessment. I t is broadly speaking true to say that the more formal the occasion, the more cxpericnced is the invited speaker likely to be and, therefore, in theory at any rate, the less in need of assistance as to the matter and manner of his address.Thcsc notes have, thercforc, been prepared primarily for those about to “read” tlieir first papers at scjcntific meetings. The qualifying v7ord “primarily” is used both becausc it is certain that even some of tlie more experienced could derive benefit from advice and 1)ecause there is just a hope that they may. The “quotes” round the word “read” are used for reasons that will soon become apparent. The first thing to do, then, is to consider just why papcrs, that is, descriptions of an author’s original work (research), are communicated to meetings at all. Clearly not as a primary means of telling every one concerned about his results, for this will be in every way better achievcd by the printed publication for wide disscmination of his finding in a scientific journal: also clearly not as a way of telling his niore immediate colleagues what a clever fellow he has been, for to do this must be either superfluous or immodest.The object of these communications is not even to give brief advancc notice of results that may save others from wasting their time by repeating what thc author has already clone or by using a technique less effective than his. That particular ohjcct is bctter, because much more quickIy, achieved by submitting this brief advance notice to a scientific journal, for which purpose arrangements for rapid publication arc now made in several journals of repute. No, the young, and the old, author should have one primary and one secondary object in view, and no others, if lie decides to “communicate.” The primary objective is consulta- tion with colleagues.He wants to tell them what lie has done and wliat he thinks his results mean. He wants them to tell him if, in tlieir opinion, he Is about to make an ass of himself in trying to publish his resuIts or his conclusions. Unless he has sufficicnt humility to envisage this impleasant possibility, he had far, far better keep his mouth shut. But if he honestly seeks constructive criticism-and he can scarcely lay claim to the term “scientific” unless he does-then hc has an immediately corollary secondary objective. And that is to interest his colleagues sufficiently to induce them, having heard him and having something to say, to get up and say it.From object nurnbcr one it must follow that it is not as a rule worth while t o make a. scientific “communication” of a completed and accepted paper. Unless the author is prepared to put himself, or the publishing society, to unwarranted expense, there is little point in offering for criticism material so far advanced as to have been submitted for publica- tion in final form. The journal’s editors and refcrces will look after the criticism business at that stage, anyhow. In our Society, it is true, authors of accepted papers are sometimes invited to present them at ii meeting bcforc publication. It might be better if these authors came forward at an earlier stage and offered to communicate their results to their colleagucs before makjng any attempt to put them into final shape.In either event, the material should in no circumstances be presented as if it were all ready for publication, that is to say, it should nevcr be solemnly rcad to an audience, word for word, as it exists in the typescript submitted to the editor. ‘To do this is not only insulting to the audience, for it is clearly wasting their time if an author reads to them something about which he has obviously alrcady made up his mind, but it is almost one hundred per cent. certain to bore them. Reading from a script in a manner to interest the listener is one of the most difficult of tasks-as the B.B.C. knows full well, Reading verbatim from typescript is a well-tried procedure guaranteed to make it impossible to do either of two othenvise perfectly com- patible things, to interest an audience and sincerely t o a5k for its members’ honest criticism.Thcrc are many ways of communicating research results without recourse to a typed script, and every research worker must ultimately work out his own detailed technique. This admittedly is a pis alley.Sept., 19541 COMMUNICATIONS AT MEETINGS I N GENERAL 531 If he really knows his subject-as he must do if he has actively taken part in, and not merely “directed,” the research-if he is himself interested in it-and he should not be presenting it unless he is-and if he is honestly seeking the help of his colleagues-the only real justifica- tion for his presence on the rostrum-then the other desired things will nearly always follow. But if any one of the essential conditions is lacking, he runs a risk, serious almost to the point of certainty, of not only being a bore, but also of appearing a conceited bore! If then the “communicant” is to do without a typescript, and if he is, by hypothesis, somewhat of a novice and correspondingly nervous, he needs something to lean on besides his memory. Even though this should be exceptionally good, it is liable to let its owner down just at those moments of crisis when he is ditheringly facing an audience he respects.Let him, therefore, decide first exactly which of his experimental results he wants to discuss. These must be arranged, condensed and tabulated in such a way as to form the basis of good clear lantern slides. The tables are by no means necessarily the same as those he will wish to see published in his printed paper: some may, indeed, be devised solely for the purpose of his “lecture,” to condense conveniently matters described at more length in the text. Whether the slides are designed ad hoc, are condensed versions of tables intended for printing or are identical with these, they must in no circumstances whatever (for the usual 3i-inch x 3i-inch slide and normal projection conditions) contain more than 12 rows (horizontal) or 10 columns (vertical) of figures in tabular form and will be far more effective if they contain from a quarter to a half of the corresponding number of entries (30 to 60 separate figures). Slides of descriptive, as distinct from numerical, matter should similarly be restricted in content.Not more than six (or perhaps eight) lines, with plenty of space between lines, and not more than eight or ten words per line, can possibly be taken in by an audience: any one who thinks differently is himself being taken in by his own close familiarity with the subject described.Graphs, again, should be bold, with figures and letters for the co-ordinates clearly indicated, and it is utterly useless to throw on the screen a slide showing several curves running meaningless neck-and-neck races with each other. Remember that there are other graphical devices besides graphs : histograms and variegated columns (vertical or horizontal) will often drive your point home better, that is, more quickly on the screen, even though curves are adjudged more accurate and otherwise informative for the finally printed page.Best of all these are made in Indian ink on white card (or a white Perspex plate) by means of stencils: next best is to have some one gifted with a good calligrapher’s script to draw the diagram, using the same media. As a last resort they may be typewritten, provided only capital letters are used and punctuation marks kept to a minimum; a new very dark ribbon is essential and it is a good trick, if the typing has to be done on paper (not card), to use a carbon paper with the carbon side uppermost behind the sheet to be photographed. Card, plate or sheet is then converted via negative to lantern slide in such a way as to secure maximum contrast. Test your slides, whenever humanly possible, on the projector and in the room where the communication is to be made; if this is out of the question, do the test under conditions as near as possible to these.The slides being actually or potentially in existence as the backbone of your discourse, you now have to decide on its flesh and clothing. The simplest procedure is the one most usually adopted, unfortunately, and the one to which undeviating hostility has already been expressed. That is, to read out word for word all or part of a text prepared for publication as a printed paper. There is no objection to having such a text with you, for reference during such discussion as your presenta- tion may arouse, but if that provides you with the slightest temptation to read instead of talk to your audience, you had better have left it at home.Best of all, though there are few novices who are brave enough to try this-far fewer, we believe, than the number that could do it if they tried-is to talk about your carefully planned and properly arranged slides without any notes at all. A series of cards, each bearing a print corresponding with one of your slides, and arranged in the same order, is a most useful aide memoire, particularly if your slides number more than three or four. Such cards, with very little practice, will enable most investigators to talk about their work to their audience (instead of to the lantern screen, as they will have the facsimile of the projected image between themselves and the audience) and in the impromptu and spontaneous manner most likely to encourage comment and criticism. The closest possible attention must be paid to letters and figures on slides. You should never do this in any circumstances whatever.532 COMMUSICATIOSS AT MEETINGS IX GEXERAL p o l .79 Those who are not prepared, at any rate on the first few occasions, to rely solely on “picture-cards” of the kind described can modify it slightly so as to have a series of “pointer” notes under their very noses, yet without any risk of beginning to “read out loud” to the audience. If the prints used are made from the same negatives as are the lantern slides and are then mounted on plain postcards, there will be enough space to one side or the other for the inscription of a few key-words in bold capitals. These should enable all but the most temerarious to talk without other written text 10 a friendly audience for 10 to 30 minutes- and no audience should be expected to remain friendly to any research worker who talks to them for more than 30 minutes about one of his investigations.Indeed, it may even be said that any author who finds, or thinks, he cannot face an audience without something more than this amount of mnemonic material in front of him ought to give up making oral presentations until he can. One further word of warning about lantern slides-for it must now be realised that they constitute the bed-rock foundation of the good “communication.” They must have been checked and re-checked for serial order: nothing is more discourteous to an audience than to have to tell the operator (and incidentally nothing is more unfair to the operator) that he is showing a slide out of order.When he does, it is always the author’s fault. Equally unfortunate is the appearance of an image on the screen sideways, upside down or back to front. This is also always the fault of the author for not having properly spotted the slide. How to do so correctly has been laid down by the British Standards Institution: British Standard Specification No. 1917 : 1952, “Film Strips and Lantern Slides.’’ This requires a “thumb-spot” (a circular or square white disc) in the left-hand bottom corner on the front of the slide (that is, on the same side as the slide’:s$Zm surface). I t will be put by the operator on the screen side of the slide carrier and at the top left-hand corner as he looks towards the screen.There is no objection to having also a white strip along the top front of the slide (this is convenient, anyhow, for writing in a reference number or a brief description), but the thumb-spot is the essence of the lantern slide contract and must in no circumstances be omitted. Naturally, it should for permanence and security be inserted between slide and cover-glass when these are bound together. Remember, there is only one right way of inserting a slide in the carrier, but there are seven wrong ones! What we have so far discussed here are mainly mechanical contrivances, in particular slides, texts and notes. I t is necessary now to turn to the much more elusive subject of delivery and manner. Being so much more elusive it will have, perforce but perhaps fortunately, to be treated much more briefly.I t may not be possible to turn everyone into a first-class speaker, but there is hardly ,anyone (at any rate, hardly any intelligent scientist who is master of his subject) who cannot with persistence and humility be prevented from being a bad or even a poor speaker. He may, however, have carefully to learn certain things before he ought to be let loose on any platform. First, look at the audience, as many of thejm in turn as possible, but always at some of them. They are not interested in a view of the top of your head, or even in an oblique view of your forehead, especially if they cannot hear what you are saying anyhow. And, therefore, secondly, speak up in such il way that the rear-seated audience can hear you. In fact, address them and not the people in the front row-not even the chairman.Remember too, to keep your tone of voice up as far as the end of each sentence. Keep your sentences reasonably short. If you want to qualify a statement, do not do it by inserting a parenthesis in the middle of a sentence. This may easily get you (and the audience) hopelessly lost. Your slide photograph (or note) should be in front of you to enable you to get back to the main line of argument. Incidentally, card notes (and photographs) of the kind suggested have the advantage that, unlike type- written or handwritten documents, they can be held in the hand of a speaker who remains vertical and looks across at his audience (instead of up to them) while he is addressing them. The planners of the meeting will have allotted to each communication and the discussion on it a certain portion of the meeting’s whole time. They will, either broadly or precisely, have divided your particular allocation into time for communication and time for discussion.Every additional minute you take for the former will have to come off the latter. Should you, indeed-as I have, alas, known happen more than once-spend the whole of your allotted time in presentation, the Chairman will risk sharing your deserved unpopularity unless he thereupon guillotines Of one general truth there is no question. Qualify in a separate sentence. Then there is the over-riding question of time keeping.Sept., 19541 THE PRESESTATIOS OF PAPERS AT MEETIXGS OF THE SOCIETY 533 all discussion. In that event, if you accept the main assumption of our thesis, that you have brought your results to the meeting to be discussed and criticised, vou will have been guilty of an unpardonable act of extreme dmourtesy, for you will have wasted the time of everyone present (and incidentally your own also).Apart, however, from the obvious turpitude of exceeding the time allotted t o you for presenting your results, there is everything to be said for taking even less time than you have been allowed. There is nothing-not the most brilliant scientific discovery, not the most scintillating comments on his own and others’ work-that so permanently endears an author to his audience, makes it certain that his reappearances on the platform will be welcome and applauded, and puts his name high among those venerated by the meeting’s organisers, as does his successfully under-running his time. Instead of a sigh of relief from the audience when he sits down, he is likely to elicit a round of cheers. The man who takes 9 minutes to present a communication for which he has been allowed 10 will be the uncrowned king of the evening’s proceedings. It can be done, but it needs foresight, the acceptance of hard work and an inflexible will. None of these qualities, be it noted, have anything whatever to do with ability as a speaker. This you may not possess now, or ever, but the others are as likely to be your characteristics as they are to be those of any other chemist. Why not use them? A. L. BACHARACH

ISSN:0003-2654    

DOI:10.1039/AN9547900530

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Sept., 19541 THE PRESESTATIOS OF PAPERS AT MEETIXGS OF THE SOCIETY 533 The Presentation of Papers at Meetings of the Society A LITERAL interpretation of the common colloquial expression “to read a paper” is a frequent cause of disappointment to the inexperienced author and to his audience. To listen to a scientific paper that has been written for the eye of a reader, in the cold, colouriess words proper to a scientific subject, is a monotonous and boring experience for the kindliest of audiences; and most scientific audiences are kindly, even when they are almost too bored to keep awake. The trouble lies in the difference in style and diction between what is suitable and necessary for the eye and what is appropriate and essential for the ear. A further complication is the destruction of that bond of sympathy, which should exist between a speaker and his listeners, by the conscientious author feeling himself under the obligation to account for every word of his written script.This can only be done with a bowed head and averted eye. Nobody listens with pleasure to one who turns away his eye and, at the same time, all too often renders himself inaudible to those beyond the range of a few yards. By doing this, the lecturer creates for himself the major difficulty of the broadcaster, without any knowledge of the broadcaster’s technique for overcoming it, or providing any mechanism by which the listener can “turn him up” or “turn him off.’’ For his own sake as well of that of his audience, an author should never “read apaper” in the literal sense.He should face his audience and talk of his work as though telling a few chemical friends the story of what he did, why he did it, and the results he got. Should he require an aide-memoire, a series of cards similar to that described later as bearing copies of the lantern slides-or the same series for both-can be prepared ; preferably they should contain only the headings of the subjects to be talked about. The short formula for this method of presentation is the same as for a successful after- dinner speech: tell your listeners what you are going to tell them-tell it to them-and then tell them you have told them. In practice, this becomes ( a ) the preamble or introduction, giving the reasons that led to the work recorded in the paper and describing, in outline, any previous work on the same subject, (b) the main subject of the paper and (c) the peroration or closing remarks that bring the discourse to a pleasing and natural end.In the introductory part it is not advisable to give full references to all previous workers in the same field; a concise account of their work, with perhaps brief mention of the more important names, is as much as an.audience can be expected to retain. A statement that full references will appear in any printed paper can be made, if thought desirable, but is hardly necessary in view of The Awdyst’s well-known custom and style. After the introduction, it is appropriate to give a short account of such experimental work as may have been a necessary preliminary to the main subject of the paper.534 THE PRESENTATION OF PAPERS AT MEETINGS OF THE SOCIETY In describing original work, greater detail is both permissible and desirable, but even at this point, tiresome detail and minutiae of manipulation should be avoided so far as possible and only given where they are essential to a proper understanding of an analytical procedure.Some results on synthetic mixtures, when the nature of the subject makes them possible, are very desirable; these carry conviction to an audience more completely than mere verbal statements on accuracy. If the paper describes a method likely to be of commercial importance, a short table of results for a range of commercial products is essential. The presentation is best brought to an end by a concise recapitulation of what has been accomplished and mention of any special advantages that the proposed method may possess over methods already known.During the course of the lecture the fullest use should be made of lantern slides and diagrams. These help to hold the attention of an audience and are the most rapid means of conveying information. The preparation of slides is well worth the effort involved, as it takes time to draw even the simplest diagram on the blackboard, and this interrupts the lecture. The slides used at the presentation of a paper may well be more numerous than the illustrations required (or permitted) in the printed version of the paper. Lantern slides must not carry more detail xhan can be seen at a glance; they should be designed to emphasise the main points of the paper.They need not necessarily carry as much information as is desirable in the paper, but they should be carefully prepared from clear diagrams or, if they are photographs, should be the best of their kind. Tables should be drawn up with the aid of stencils and photographed and a black-on-white slide should be made: negative slides, with white lettering on Ellack, are difficult to read from any distance, and it is impossible to see a pointer when it is used on them. It is no compliment to an audience to show it blotched drawings or tables liurriedly made on the gelatin of a fixed plate, or photographs that lack detail and require verbal explanations of what should be clear to the eye. A slide held at arm’s length against a. white background should be readily legible.Only in an emergency should the “reflecting” half of an epidiascope be used-it is not as satisfactory as properly prepared slides. If the episcope is used, tables and diagrams should be drawn in black ink (typewritten matter is too small), and the drawings and tables must be kept within the area available-usually 6 inches x 6 inches. Again, the set of drawings must be clearly numbered, or confusion will ensue. Illustrations should not be left on the screen after the lecturer has passed on to another part of his subject; an audience left in the dark to gaze on pictures that no longer have any connection with what is being said soon loses interest. A few words like “so much for that” or “we will now pass on to . . .” will tell the lanternkt that the slide is no longer wanted and he will switch on the lights.The lecturer should provide himself with a list of his slides, and drawings for the epidiascope, in the order in which he will require them. A series of cards, each bearing a print made from the same negative as the corresponding slide, is a most useful aid. Slides should be numbered and arranged in the order in which they will be wanted. After the slides have been arranged in order, it is advisable to seal the box and resist all requests from friends to be allowed to see them. Nobody should be allowed to touch them but the lanternist, Nothing can be more disconcerting to a lecturer than a batch of slides that come through in the wrong order. Also, as there are seven wrong ways of inserting a slide in the projector and only one right way, it is essential that each slide should bear a “thumb-spot” in the left-hand bottom corner on the front (film) surface: this has been laid down in British Standard No. 1917 : 1952 “Film Strips and Lantern Slides.” A meeting of the Society usually lasts about two hours, during which time three papers are presented and discussed. This allows 40 minutes to each speaker, so that, if adequate time is to be provided for discussion, he should not speak for more than 25 minutes. It is essential for clarity not to attempt to speak at a greater rate than 150 words a minute; this will allow him about 3000 words. [Vol. 79

ISSN:0003-2654    

DOI:10.1039/AN9547900533

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Sept., 19541 WILSON 536 The Determination of Phosphate in the Presence of Soluble Silicates Application to the Analysis of Basic Slag and Fertilisers BY H. N. WILSON (Presented at the meeting of the Society on Wednesday, M a y 5th, 1954) I t is shown that soluble silicic acid, even after prolonged boiling to polymerise it, is readily converted to silicomolybdic acid by sodium molybdate in hot acid solution and is precipitated as quinolinium silicomolybdate under the same conditions as is the phosphomolybdate ; this seriously interferes with the determination of phosphate. This interference is overcome by the addition of citric acid, which forms a complex with molybdic acid of such stability that its reaction with silicic acid is prevented, whereas the reaction with phosphoric acid proceeds normally.Experimental results are quoted, and details are given for the analysis of basic slag, for which material the new method is superior to the statutory method. The use of a citric - molybdic acid reagent also simplifies the accurate determination of phosphate in fertilisers. IN 1951 a new, accurate and rapid method for the determination of phosphate was pub1ished.l In this method phosphate is precipitated as quinolinium phosphomolybdate, instead of as ammonium phosphomolybdate-a procedure much less subject to interference than the older methods. The precipitate is of the theoretical composition and hence can be titrated without further purification. As it is rapid-it takes about 2 hours-and accurate, this method has excited some interest in the fertiliser industry.It has, however, been pointed out that soluble silica will interfere to an unknown extent, and that hence there would be difficulties in applying the method to the analysis of basic slag (especially for citric acid soluble phosphate) or any mixed (“compound”) fertilisers containing basic slag, or to any other fertilisers that contain soluble silica. This paper describes an investigation into the extent of the interference of silica and gives details of a satisfactory analytical procedure that is based on a quinolinium silico- molybdate precipitation. The appendix shows that the determination of phosphate in fertilisers can be expedited and simplified. THE EXTENT OF INTERFERENCE BY SOLUBLE SILICA In the volumetric determination of silica by precipitation of quinolinium silicomolybdate, the conditions, such as acidity, must be carefully controlled if the results are to be quantitative.2 Also, in the colorimetric determination of silica as silicomolybdic acid, care is always taken to adjust the pH to the optimum value for the rapid formation of silico- molybdic acid.There is some doubt as to the best pH value; for example, Stross3 prefers pH values of 0.7 to 0.8, the American Public Health Association Standard Methods for Examination of Watel-4 requires a pH value of 1.2, and it was thought probable that, at high acidities, formation of the complex acid would be slow and its effect would be small. Further, it is usually considered that the silica must be in true solution and must not be polymerised or colloidal; e g ., the A.P.H.A. method4 specifically refers to “crystalloidal or non-colloidal silica.” (Although in this connection “crystalloidal” has a vague and uncertain meaning, that of “non-colloidal” is clear enough.) Nevertheless, Malaprade says “silico- molybdic acid is formed almost instantaneously when one causes a colloidal solution of silica to react with a solution of MoO,.~H,O”~ or “by acidification of a mixture of silicate and molybdate. We have been able to extend the application of the use of molybdic acid as a reagent for free silica whether colloidal or not. Gelatinous silica, it is true, reacts only slowly on solutions of Mo0,.2H20 in giving the yellow coloration.”6 These observations have been completely confirmed. At least for acidities up to 2N with respect to hydrochloric acid, silicic acid in solution, however much polymerised, is converted to silicomolybdic acid ; even the silica obtained by evaporation to dryness of an acid solution yields some silicomolybdic acid.Hence there is no possibility of preventing silica from reacting with molybdic acid by altering conditions of acidity or the state of aggregation of the silica.536 WILSON: THE DETERMINATIOK OF PHOSPHATE I N THE [Vol. 79 There appears to be a contradiction between the ready formation of silicomolybdic acid described above and the insistence on careful control of pH value, which is found in all the literature on the colorimetric determination of silica. The explanation is that there are two forms of silicomolybdic acid with the same composition but with different optical properties in s~lution.~ The p-form is the more strongly coloured and hence more suitablt for the colorimetric determination of traces of silica: it slowly changes to the a-form.The form produced depends on the ratio of hydrogen ions to molybdate ions; but conversion of silica to silicomolybdic acid (presumably the a-form) is readily effected over a large range of acidities and, when the temperature is-high, it can be made from any form of soluble silica, even if the yield is not completely quantitative. USE OF CITRIC ACID TO OVERCOME INTERFERENCE BY SILICA It is fairly well known that soluble silica also interferes with the colorimetric determination of phosphate as phosphomolybdic acid, and that this interference can be overcome by the use of tartaric or citric acids: e.g., Muirs recommends the use of tartaric acid in soil analysis, and Zimmermann,g in a brief article, discusses both citric and tartaric acids and points out that too much hydroxy acid will also inhibit the phosphate coloration.The mechanism of the reaction is obscure, and although the facts are not in dispute, the action of citric acid or certain other acids on molybdic acid is quite complicated. If the citric acid is mixed first with the molybdate in acid solution and then added to the acid solution of the phosphate and silicate, only the phosphate reacts, but Malaprade says: “Certain substances that in the cold give molybdic complexes, e.g., periodic, nialic and tartaric acids, or mannitol, destroy phosphomolybdic acid, and under similar conditions do not react with silicomolybdic acid.”lO This suggests that silicomolybdic acid is more sta.ble towards such reagents than is phospho- molybdic acid, but under other conditions the reverse is true.Malaprade continues: “Experiment has shown us that to detect silica. as silicomolybdic acid one must add the reagents in the proper order. We give here several experimental facts which we cannot explain. . . . Under these conditions, tartaric or periodic acid have no action on silicomolybdic acid already formed, but if one tries to make silica react with a previously prepared mixture of molybdic acid and tartaric or periodic acid, there is no formation of silicomolybdic acid, even after several days, even in the hot solution.”1° What is true of tartaric acid is true of citric acid, but the previously prepared mixture of citric and molybdic acids does react with phosphate in acid solution.Whilst the mechanism of these reactions between citric and molybdic acids a t different pH values and under different conditions is a fascinating subject for chemical speculation, in default of more experimental results it is better to refrain. Suffice it to say that the facts are in accordance with the hypothesis that in acid solution, hot or cold, silicomolybdic acid is less stable than phosphomolybdic acid, and that by addition of citric acid to solutions of either molybdic acid or sodium molybdate containing free hydro- chloric acid, the amount of free MOO, present ca.n be controlled; hence a ratio of molybdic acid to citric acid can be found such that the mixture will form phosphomolybdic acid and not silicomolybdic acid.. The exact constitution of the molecular or ionic species in acid molybdate solutions is uncertain: numerous species have been described; some of them no doubt exist.Most if not all polymolybdic acids can, however, be represented as H,MoO,.rtMoO,, and, as the heteropoly acids all contain MOO,, we may represent these complicated reactions in a simplified form by regarding MOO,” and MOO, as the species concerned. Then the (no doubt over-simplified) equilibria may be written as below, and will at least qualitatively represent the course of the reactions- [H,SiO,] [MoOJ” [H,SiO,. 1 BMoCG] [H,PO,I [MOO3 ]” [H,POg. 12Mo0,] Kl = K , = [MoO,”][H]~ K3 = [MOO31 [molybdic anh:ydride] [citric acid] [citric - molybdic acid complex] ’ K4 =Sept., 19541 PRESENCE OF SOLUBLE SILICATES 537 and in strongly acid solution K , is larger than K,.As the molybdic anhydride concentration is raised to its nth power (although, in all, 12 moles of MOO, combine with 1 mole of H,SiO, or H,PO,, n will certainly not be as great as 12, but may well be greater than 1), changes of citric acid concentration will have a related effect on the amount of silicomolybdic acid that will be formed, and if the quantity of citric acid is too great, the formation of phospho- molybdic acid will also be hindered or prevented. As the MOO, concentration also varies as the square of the hydrogen-ion concentration, changes in acidity will affect the precipitating power of the solution.In the presence of given amounts of citric acid and molybdic acid, if the concentration of hydrochloric acid is too low, phosphate is incompletely precipitated. There is also an upper limit to the acidity that can be tolerated. Hence the concentration of “available” MOO, can be so controlled that the solubility product of quinolinium silicomolybdate is not exceeded, whilst quinolinium phosphomolybdate is precipitated quantitatively. The experimental results show that the .range of permissible concentrations is great enough for precise determinations of phosphate to be made in presence of silica. EXPERIMENTAL This is confirmed by experiment. EFFECT OF CITRIC ACID ON PRECIPITATION O F QUINOLINIUM SILICOMOLYBDATE- It was found that in hot acid solution, tartaric acid caused some reduction of the molybdate, precipitates being blue or bright green instead of yellow, so citric acid was used.A solution was made by fusing 0.25 g of pure quartz with 2.5 g of sodium carbonate, dissolving the melt in water containing 2 g of sodium hydroxide and diluting it to 500 ml; 25 ml of this solution, containing 12.5 mg of silica, were diluted to 100 ml with water, and 20 ml of concentrated hydrochloric acid and then 30 ml of 15 per cent. sodium molybdate solution were added. The whole was heated to boiling point (intensification of the yellow colour occurred as the temperature rose), and the quinolinium salt was precipitated by slow addition of an excess of 2 per cent. quinoline solution.After 15 minutes the flask was cooled to 15” C, the precipitate was filtered off and washed, and titrated by the procedure for the deter- mination of ph0sphate.l The experiment was repeated in the presence of increasing amounts of citric acid, and a blank determination in presence of citric acid was also performed. The value of this blank, which must be due to the presence of phosphate (not silicate) in the reagents, was deducted from each determination; it amounted to 0.15ml of 0.5N sodium hydroxide solution. The results were- Citric acid present, g . . .. 0 0.5 1.0 2.0 3.0 4.0 5.0 Silica determined, mg . . . . 11.8 8.1 3.5 0.9 0.4 0.1 nil Hence it is possible, by the addition of citric acid, to prevent silica from reacting, and although a reagent of different composition from that used previously1 is finally recom- mended, it is possible to deal with minor amounts of silica simply by adding citric acid to the acid solution before adding the sodium molybdate reagent.Addition of 5 g of citric acid led to results for P,O, that tended to be low (average of six experiments, 10.99 per cent. ; known content, 11.04 per cent. of P,O,), so it was decided to vary the composition of the reagent systematically to arrive at the optimum conditions. These experiments gave results con- sistent with the equilibria suggested above. VARIATION I N COMPOSITION OF REAGENT- Efect of excess of molybdate-The precipitation was carried out in 150 ml of solution containing 20 ml of concentrated hydrochloric acid, 5 g of citric acid and 12.5 mg of silica in solution. The amount of 15 per cent.sodium molybdate solution added was varied, and the silicomolybdic acid was precipitated with quinoline and titrated by the usual method. The results were- Volume of molybdate solution Silica precipitated, mg . . .. 0.10 0.6 1.7 8.0 12.0 12.65. added, ml . . .. .. 30 40 50 60 70 90 Hence an increase in the ratio of molybdic acid to citric acid is shown to be undesirable. Efect of change in mineral acid concentration-Changes in the concentration of mineral acid have not been found greatly to affect the formation of silicomolybdic acid but, in the presence of citric acid, the mineral acid concentration is an important factor in the formation of phosphomolybdic acid. In the absence of mineral acid but in presence of 5 g of citric acid538 WILSON: THE DETERMINATIOli OF PHOSPHATE IN THE [Vol.79 there is no formation of phosphomolybdic acid, but on the dropwise addition of hydrochloric acid the yellow colour appears and deepens with incre--’-.., dmounts of acid. A series of aliquots of a fertiliser solution (P205 content of fertiliser, 11.06 per cent.) were treated as in the previous methodl for the removal of ammonia with bromine, The solutions were made just acid with hydrochloric acid, boiled to remove bromine, the volume was adjusted to about 125 ml, and 5 g of citric acid and various amounts of hydrochloric acid were added to each. Thirty millilitres of sodium molybdate solution (equal to 4-5 g of solid) were added to the test solution, which was then boiled, and the phosphomolybdate was precipitated. Three of the test solutions contained 12.5mg of silica.The results of the determinations are shown in Table I. TABLE I EFFECT OF VARIOUS CONCENTRATIC~NS OF MINERAL ACID ON THE DETERMINATION OF P,05 Concentrated hydrochloric acid added, ml 1 2 3 4 5 10 15 5 10 16 20 Silica added, P,O, found, mg % - 2-79 - 8.20 - 10.83 - 11-01 - 11.04 - 11.04 - 11-04 12.6 11-04 12.5 11-05 12.6 11.05 - 10.89 Although in the absence of citric acid the presence of 20 ml of hydrochloric acid is usual and even 25 ml does no harm, it can be seen from the results that the hydrochloric acid concentration must be lower in the presence of citric acid. Whether this effect is caused by the citrate ion or by the higher acidity was not ascertained: the “strength” of citric acid as an acid will be increased by complex formation, and hence it will make a significant addition to the hydrogen-ion concentration.Efect of change in citric acid concentration--At about this stage in the investigation, difficulty was experienced in obtaining sodium molybdate of good quality. Traces of silica as an impurity, although no longer important in analysis-its effect could be nullified by the addition of a little citric acid-were undesirable in rtn investigation ; some supplies of molybdate also contained phosphate. It was found that a stable solution could be made by dissolving molybdic acid in a solution containing citric and hydrochloric acids , and several concentrations were tried. The first solution that was used in this series of experiments consisted of 3.4 g of molybdic acid (equivalent to 4.5 g of Na,Mo0,.2H20), 5-0 g of citric acid and 10.0 ml of concentrated hydrochloric acid made up to !?Om1 with water.A new solution containing silica was made b,y diluting a weighed amount of “silicate of soda,” previously analysed, pouring it into 15 11-11 of concentrated hydrochloric acid and diluting it so that 1 ml of the solution contained 20 mg of silica. Various volumes of solution were diluted to about 120 ml, 50 ml of the above reagent were added, and the precipitated silica was determined in the usual way. The results were- Silica present, mg . . .. .. 25 50 100 200 500 Silica found, mg . . .. .. 0.6 0.9 1.1 1.1 1.5 A further 2 g of citric acid was added to the reagent so that it contained 7 g of citric acid.With this reagent, from a solution of 250 mg of silica only 0.4 mg was precipitated. Similarly, with a reagent containing 9 g of citric acid, the titration value was equivalent to 0.1 mg of silica from 500 mg of silica-possibly a blank, and not even silica. On the basis of the above work, a number of reagents were prepared and tried on known mixtures of silicate and phosphate solution. (The investigation was unduly prolonged because the material purchased as “molybdic acid” was found not to be molybdic acid at all, but to approximate to (NH4),Mo04.3Mo0,, and the author was not aware of this until half-way through the experimental work. The chemistry of the molybdic acids is far fromSept., 19541 PRESEKCE OF SOLUBLE SILICATES 539 simple, and it is unfortunate that it should be made to appear even more complicated by the perpetuation of a name that exhibits the mistaken view of a previous generation.) I t was desirable that the reagent should be able to deal with at least twice as much silica as phosphate and that it should be readily obtainable free from phosphates, so as to avoid any “blank.” For this reason the use of sodium molybdate was abandoned, and finally AnalaR molybdic anhydride used.The presence of 7 g of citric acid would ensure that silica contents of up to 250 mg could be tolerated; as 60 mg of P,O, is the most that is ever present in an aliquot taken for analysis, this allows a comfortable margin. Amongst other variations, a reagent was prepared that contained the necessary quinoline as well as the citric and hydrochloric acids and the molybdic acid.This was added to the just acid solution of the sample. It gave a precipitate that was readily removed by filtration, and would have been quite satisfactory if it had been stable, but unfortunately it began to deposit molybdic acid after a few hours. The reagent solution finally prepared contained 5.4 g of anhydrous MOO,, 1-1 g of sodium hydroxide, 14 g of citric acid and 12 ml of hydrochloric acid, sp.gr. 1-18, made up to 100 ml with water. Fifty millilitres of this reagent (see procedure) were added to a neutral or just acid solution of the sample. Reszdfs with new reagent-Three series of experiments were made with known amounts of P,O, (from recrystallised potassium dihydrogen phosphate) and silica (soluble silicic acid), the P,O, being determined as described below.Results were as follows- -4. B. C. 50mg OF P2O5 PRESENT. VARIOUS AMOUNTS OF SILICA- SiO, present, mg . . nil 60 100 200 500 750 1000 P205 found, mg . . . . 49.89 49.86 49-96 49-93 {4g.93 49-86 50.06 49‘93 { 5046 200mg OF SILICA PRESENT. P,O, present, mg . . 10 20 30 40 50 60 70 P205 found, mg . . . . 10.86 20.40 30.26 40.16 49-93 59.97 69.73 VARIOUS AMOUNTS OF P205- 50mg OF SILICA PRESENT. VARIOUS AMOUNTS OF P2O5-- P205 present, mg . . 5 10 20 30 40 50 60 P205 found, mg . . .. 5-7 10.4 20.01 30.05 39.96 49.89 59.76 DISCUSSIOX- It is admitted that the results under A tend to be low (the negative error is 0.18 per cent.), which may suggest that the potassium dihydrogen phosphate was not quite pure or that less citric acid should be used.But it must be remembered that we are dealing with equilibria, and it is difficult to inhibit the effect of silica completely without some slight effect on the formation of phosphomolybdic acid. Results under B and C do suggest that there is a slight compensation of errors. This investigation was undertaken primarily, however, to find a means of applying the quinolinium phosphomolybdate method to the analysis of basic slag, which also contains iron, aluminium and manganese. These elements also form complexes with citric acid, and introduce a further set of equilibrium reactions that tend to diminish the availability of the citric acid. APPLICATION TO THE AKALYSIS OF BASIC SLAG Numerous analyses of basic-slag samples have been carried out for total and citric acid soluble P,O, both by the new method and by the statutory method; they are shown in Table 11.In addition, in many analyses an aliquot from the total P 0 solution prepared according to the statutory method was examined for P,O, by the "aid" quinoline method, which is regarded as more accurate than the statutory gravimetric procedure. These analyses show that the composition of the new reagent is satisfactory. Many of the magnesium ammonium phosphate precipitates, both from total and citric acid soluble P203 determinations contain enough iron to make the magnesium pyrophosphate cream coloured instead of white. Perhaps this iron is carried forward as a ferrimolybdate complex. I t will cause the “official” results to tend to be high.THE INSOLUBLE MATTER IN BASIC SLA4G- In the statutory method for the determination of total P,O, i,n basic slag, efforts are made to separate as much silica as possible by evaporation to dryness-or, more often,540 WILSON: THE DETERMINATION OF PHOSPHATE I N THE [Vol. 79 pastiness-of the acid solution. These efforts are only imperfectly successful ; some milligrams of silica always remain in solution and may contribute to the tendency towards high results. Alternatively, the large insoluble residue (mostly dehydrated silica) obstinately retains some phosphate. This is recognised in the Statutor!y Regulations, which state : “The insoluble matter is to be washed from the filter, re-extracted with acid and any phosphoric acid present in the solution added to the main quantity”-an ambiguous and badly drafted sentence.To effect complete recovery of this residual phosphate enmeshed in the gelatinous silica, it is necessary to wash the insoluble matter off the paper back into the beaker, boil it for some time with hydrochloric acid, again evaporate the solution to dryness, extract the residue with dilute acid, and again filter and wash it. ’This time-consuming process recovers on an average 0.45 per cent. of P20,, but the amount depends on the sample; the following results show the amount recovered for six samples- Sample number . . .. .. 5 6 7 8 9 10 Recovery of P,O,, per cent. .. 0-56 0.41 0.76 0.42 0.46 0.28 In the new procedure this laborious process is avoided. The aim of the new method of extraction is to dissolve and retain in solution all of the phosphoric anhydride and all the silica, and so avoid the troublesome second extraction.This is achieved by mixing the sample, powdered to pass through a 100-mesh sieve, with a large volume of water, warming the slurry and adding hydrochloric acid whilst stirring. Except for a small residue-mainly iron and manganese oxides-the whole sample dissolves to form a stable solution. Ferrous iron is oxidised with nitric acid, the solution is diluted to known volume, filtered through a dry No. 30 filter-paper and the determination is made on an aliquot of the filtrate. The insoluble matter from six samples has been weighed and examined for P,05; each sample weighed 2.5 g and yielded insoluble residues as follows- Weight of residue, mg .. . . 111.4 82.8 133 57 . Sample number . . .. * . 1 3 5 9 13 15 P,O, in residues, mg .. .. 2.7 0.1 1 0.63 0.72 0.17 0.21 P,O, undissolved (calculated on 68.1 33-6 original sample), per cent. . . 0.11 0.004 0.025 0.03 0.007 0.008 Such small amounts as these can be considered negligible, but should anyone wish to recover these minute quantities, it is far less trouble to wash and ignite a small residue, fuse it with a pinch of sodium carbonate, dissolve the melt and add it to the original solution than to extract a gram or so of gelatinous silica. As a matter of possible interest, we have determined the quantity of soluble silica present in the usual aliquot taken for P205 determination. It ranges from 60 to 70mg. METHOD FOR ANALYSIS OF BASIC SLAG SPECIAL REAGENTS- Citric - molybdic acid reagent-Stir 54 g of AnalaR molybdic anhydride, MOO,, with 200 ml of water, add 11 g of AnalaR sodium hydroxide and stir it, whilst heating, until the molybdic anhydride dissolves.Dissolve 120 g of pure citric acid in about 250 to 300 ml of water and add 140 ml of pure hydrochloric acid, sp.gr. 1-18. Pour the molybdate solution into the acid solution, which is stirred throughout the addition. Then cool it and filter it through a -pulp-pad, if necessary. Dilute the solution to 1 litre. This solution is slightly green or blue in- colour and the colour deepens on exposure to light. Add dropwise a dilute (0-5 or 1.0 per cent.) solution of potassium bromate until the green colour is discharged. Quinoline solution-Measure 60 ml of concentrated hydrochloric acid and 300 to 400 ml of water into a l-litre beaker and warm it to 70“ or 80” C.Pour 50 ml of “synthetic” pure quinoline (free from reducing agents) in a thin stream into the dilute acid, whilst stirring. When the quinoline has dissolved, cool the solution, dilute it to 1 litre and filter it through pulp. Synthetic quinoline is usually suitable, but a satisfactory method of purifying impure quinoline is to dissolve it in hydrochloric acid and, with an excess of zinc chloride dissolved in dilute hydrochloric acid, to precipitate the double chloride, (C,H,N),ZnCl,. This crystallises well, and is washed with cold dilute acid. The quinoline is regenerated with an excess of sodium hydroxide solution, dried and then distilled.Indicator soZ.ution-Mix 3 volumes of a 0.1 per cent. solution of thymol blue (dissolve 0.1 g in 2.2 ml of 0.1 hr sodium hydroxide and 50 ml of industrial methylated spirits and This will tak:e 20 to 30 minutes. This solution is now stable, if kept in the dark.Sept., 19541 PRESENCE OF SOLUBLE SILICATES 541 dilute it t o 100 ml with water) with 2 volumes of a 0.1 per cent. solution of phenolphthalein in 60 per cent. ethanol. TABLE I1 RESULTS FOR BASIC SLAG AND PHOSPHATE ROCK Sample number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17* 18* 19* Total P,O, A I > Statutory method, 11.78 % i I:;} 16.14 16-11 16.08 8.39 9.15 9-14} 10.45 10.24 14-55 15-94 19.64 19-22 i%} ;;:it} 19.68 18-97 33-20 33.43 33.47 33.0 33-53 33-46 Average for samples 1 to 16 14-94 Average for samples 17 to 19 33-33 Old quinoline method 11.62 % ; i::; } 16.06 19-23 19.01 17-68 17-68 17-65 18.88 I- iE} 18-71 Citric - molybdic acid method, 11.71 11-72 11-80 11.79 11-69 % 16.05 16-28 16.27 8-23 9.10 8.99 10.22 10-28 14.40 15.81 15.89 19-42 19.29 18-03 17.95 17.92 19-10 19.88 19.82 18.86 32-89 32.91 33.46 33.46 33.39 33.41 14.90 33.26 Citric acid soluble P,O, Statutory method, % 10.44 1 10.88 10.92 10.65 15-41 15.38 15.41 6.94 7.51 9.02 9.38 11.27 14.13 14.15) 16.92 17-02 16-88 16.82 17.32 16.35 13-23 Citric acid mol ybdate method, 10.41 % 1 10.57 10.56 10.53 10.56 15.40 15.15 15.21 6.80 7.45 9.02 9.29 11-19 14.07 16-80 16.83 16.91 16.88 16.83 16-80 16-80 16.77 17-21 17-24 16.28 16.31 1 13.14 * Samples of Morocco rock.NOTES- Samples 5 to 10 contained particles of metal that would not pass a 100-mesh sieve.These particles were removed and weighed, and the analyses were carried out on the sifted samples. Allowance was made for the metal in calculating the final P,O, contents. Results in column 3 are for aliquots from the portion used for the statutory determinations reported in column 2. Every result in column 4 was determined on a separately weighed portion. The basic slags were analysed on 2-5-g portions, the Morocco rocks on catch weights from 1.6 to 2 g. Results in columns 5 and 6 are for aliquots from the same solution. Where results are bracketed, they were determined on aliquots from the same solution. 1. 2. 3. 4. 5. PROCEDURE FOR DETERMINING TOTAL P20b- Weigh 2.5 g of sample into a 400-ml beaker, wet the solid thoroughly with 20 to 30 ml of water and then add a further 70 ml of water with continuous stirring.Warm the mixture and add dropwise, with stirring, 10 ml of concentrated hydrochloric acid and then 5 ml of542 WILSON: THE DETERMINATIOX OF PHOSPHATE IN THE [Vol, 79 concentrated nitric acid. Gently boil the solution for 10 minutes, cool it and then dilute it to 250 ml in a calibrated flask. Filter this solution through a dry Whatman No, 30 filter- paper into a dry beaker, rejecting the first 20 to 30ml of filtrate. Place a 25-ml aliquot (50-ml for a low-grade sample) in a 500-ml conical flask fitted with a stopper and marked at 150ml. Dilute the aliquot with water to about 100m1, heat the solution almost to boiling, and then add 20 per cent. sodium hydroxide solution dropwise until there is a faint permanent precipitate, Add a few drops of hydrochloric acid to clear the solution while it is still boiling.Dilute it to 150 ml, and add 1 g of citric acid and then 50 ml of the citric - molybdic acid reagent. Heat the solution to gentle boiling for 3 minutes, From a burette slowly add 25 nil of the quinoline solution, the flask being swirled continuously. Again heat to boiling and boil gently for 1 to 2 minutes. Keep the flask hot but not boiling for a further 5 minutes to allow the precipitate to age. Cool the flask and its contents to 15" to 20" C and filter the solution with suction through a paper-pulp pad, and then wash the residue with cold water until it is free from acid. The presence of citric acid in the reagent obviates the need for dilute hydrochloric acid as the washing liquid.Transfer the paper pulp and precipitate to the original flask, rinse the funnel with water and collect the rinsings in the flask. Wipe the funnel with a small piece of damp filter-paper to ensure complete removal of the precipitate and place it in the flask. Add about 50ml of water, place the stopper in the flask and shake it thoroughly to disintegrate the filter and to disperse the precipitate, as clots of the solid are difficult t.0 dissolve in the sodium hydroxide. From a calibrated burette or pipette, add sufficient 0.5 N sodium hydroxide solution (carbonate free) to dissolve the precipitate and leave a small excess. Shake the flask to dissolve the precipitate, add 0.5 to 1-0 ml of indicator and titrate the solution with 0.5 N hydrochloric acid.The colour of the solution changes from violet to green-blue and then very sharply to yellow at the end-point. Make a blank determination on all the reagents, but use 0.1 N acid and alkali for the titration. The P,O, content is calculated from the relationship- The blank should be very small. 1 ml of 0.5 A- sodium hydroxide = 1.366 mg of P,O,. PROCEDURE FOR DETERMIXIKG CITRIC ACID SOLlJBLE P205- Extract 5 g of sample with 500 ml of cold 2 per cent. citric acid solution by shaking for 30 minutes, as described in the Statutory Regulations (Statutory Regulations, 1932, Fertilisers and Feeding Stuffs Act). Filter the solution by pouring it all at once on to a large filter, as described in the Regulations.With a pipette, transfer a 50-ml aliquot of the filtrate (or 25 ml according to the expected P,O, content) to a 500-ml flask with a mark at 150m1, and dilute it to the mark. Heat the solution to boiling, add 50 ml of the citric - molybdic acid reagent and then proceed with the precipitation, filtration and titration as described above. APPLICATION OF THE CITRIC - MOLYBDIC ACID RISAGEXT TO ASALYSIS OF FERTILISERS- Although mixed fertilisers (compound fertilisers) do not usually contain appreciable amounts of soluble silica, there are two advantages in the use of citric acid in general fertiliser analysis, but in smaller amounts than are recommended for basic slag- 1. There is no longer any necessity to take stringent precautions to avoid picking up traces of silica from glassware or reagents.This means that sodium hydroxide solution can be used after keeping it for a time in glass bottles, and flasks continue to be serviceable for much longer. 2. It is no longer necessary to destroy ammonium salts before precipitating quinolinium phosphomolybdate. Just as citric acid regulates the amount of available MOO, so that the solubility product of quinolinium silicomolybdate is not exceeded and hence the compound is not precipitated, so in exactly the same way it prevents the precipitation of ammonium phosphomolybdate, which is far less insoluble than the quinolinium salt. The amount of ammonium ion that can be tolerated depends on the concentration of citric acid: in presence of 2 g of citric acid, there is slow precipitation of ammonium phosphomolybdate from a solution containing the usual amount of hydrochloric and molybdic acids, 60 mg of phosphoric anhydride and 190 mg of ammonia, but in presence of 3 g of citric acid, no precipitation of ammonium phosphomolybdate takes place from i i s much as 300 mg of ammonia under theSept., 19541 PRESEXCE OF SOLUBLE SILICATES 543 conditions laid down; if however the solution is allowed to stand at its boiling point, after 5 minutes there will be a faint turbidity caused by the formation of ammonium phospho- molybdate.As it is unlikely that more than 300 mg of ammonia will be present in an aliquot taken for P,O, determination, we have used a reagent containing 3 g of citric acid, as this is ample for normal amounts of both silica and ammonia.The course of the analysis is simplified and expedited when there is no need to destroy ammonia. Further, as the final precipitate does not need to be washed with dilute hydrochloric acid to remove excess of molybdate, time is saved here also. A series of results found by the modified method on mixed fertilisers of various types are shown in Table 111. The last few, on Billingham C.C.F. No. 1, were determined under routine conditions. The modified method has now superseded the previously described method1 in the Billingham Research Department analytical laboratories. TABLE I11 APPLICATION OF CITRIC - MOLYBDIC ACID REAGEST TO MIXED FERTILISERS Water-soluble P,O, determined by Old molybdic Sample Potassium Statutory quinoline acid number Xitrogen, oxide, method, method, method, A r 3 Citric - % 0‘ % % % /O 1 12.0 15.0 11.06 11.05 2 5.8 7.2 1.20 1.20 3 9.7 8.9 6.37 6.37 E j 4 8.7 - 17.45 17.54 17.51 17.54 5 4-8 6.3 9.63 9.72 9.30 ;:;: / 6 12.7 13.9 11.66 11.70 11-58 7 14.0 3.80 3-75 8 12.0 15.0 12-0 1 12.04 9 12.0 15.0 12.18 12-20 10 12.0 15-0 12.53 12-53 11 12.0 15.0 12.20 12.28 12 12.0 15-0 11.79 11.82 13 12.0 15.0 12.31 12.34 Average for samples 2 to 7 .. 8.35 8.35 8.29 3.78 11.59 3.7g] Total P,O, -- Citric - Old molybdic method, method, method, Statutory quinoline acid % % % 6.36 6.20 6.84 6.88 6-90 20.08 20-10 20.24 14-90 14-86 14.88 11-75 11-71 14.26 14.28 14.23 12.37 12.34 12.45 KOTES- Samples 1 to 7 had been used in a co-operative investigation of methods of analysis in which Fison’s, S.A.I. and I.C.I. Billingham Division took part.The results quoted under statutory method or old quinoline method are all the averages of several determinations in several laboratories. Results 8 to 13 were all determined in the control laboratory of the Billingham Research Department. Some months eIapsed between the analyses, and it is possible that the low results by the new method are due to “reversion.” 1. 2 . 3. Sample 5 shows a discrepancy between the results by the old and new methods. Where results are bracketed, they were determined on aliquots from the same solution. METHOD FOR ANALYSIS OF FERTILISERS REAGENTS- acid instead of 120 g. Citric - molybdic acid reagent-Prepare it as described on p. 540, but use 60 g of citric Quinoline solution-Prepare it as described on p. 540.Indicator solution-Prepare it as described on p. 540. PROCEDURE FOR DETERMINING WATER-SOLUBLE P205- Prepare the aqueous solution exactly according to the Regulations of the Fertilisers and Feeding Stuffs Act, but avoid the use of ethanol to disperse the sample, as it may cause some reduction of molybdate at a later stage. After filtration take an aliquot that will544 WILSON: THE DETERMINATION OF PHOSPHATE I N THE [Vol. 79 contain less than 70 mg of P,O, and preferably about 50 mg, and transfer it to a stoppered conical flask marked at 150ml. Dilute the alLquot to 150m1, add 50ml of the citric- molybdic acid reagent, heat the flask to incipient ebullition and maintain it a t this temperature for 3 minutes, then heat it to the boiling point. :From a burette with a coarse jet, add 25 ml of the quinoline reagent.The reagent should be added dropwise for the first few millilitres and then in a slow stream, with constant swirling throughout to ensure a precipitate of the maximum particle size. Time spent in the precipitation will be more than saved in the filtration. Allow the flask to stand in a bath of boiling water for 5 minutes, cool it to room tempera- ture in running water, and filter the contents 011 a pulp-pad. Wash the precipitate with cold water until it is free from acid. Transfer the pulp-pad and precipitate to the original flask, using not more than 100 ml of water. Shake the flask vigorously to disperse the pulp and precipitate this is essential-add an excess of 0.5 N sodium hydroxide solution (carbonate free), shake the mixture to dissolve the precipita-te, and examine it carefully to make sure that no yellow particles remain.Then titrate the excess of alkali with 0.5 N hydrochloric acid, with mixed phenolphthalein - thymol blue indicator. 1 ml of 0.5 N sodium hydroxide solution = 1.366 mg of P,O,. Make a blank determination on all reagents, but use 0.1 N acid and alkali for the titration. The blank should be very small. Then- PROCEDURE FOR DETERMINING TOTAL P205- Weigh out 5 g of finely ground sample into a 400-ml beaker, stir it thoroughly with 100ml of water and boil the mixture. Add slowly to the boiling solution a thin stream of 10ml of concentrated hydrochloric acid and then 10ml of concentrated nitric acid. Boil it gently for 10 minutes, cool it, transfer it to a 500-ml calibrated flask, and adjust the volume to the mark.Filter the solution through a dry filter-paper, discarding the first 10 or 20 ml of filtrate, and transfer a suitable aliquot of the filtrate to a 500-ml stoppered conical flask. If the sample does not contain calcium, add 100 to 200mg of pure calcium carbonate, then sodium hydroxide until a permanent precipitate is formed. Do not use an indicator, as it is liable to be absorbed on the final precipitate and interfere with the final t itration. Dissolve the precipitate by dropwise addition of dilute hydrochloric acid, dilute the solution to 150 ml with water and proceed as described above. The author wishes to thank Mr. R. Donald (Messrs. S.A.I. Ltd.), Mr. K. A. Sherwin (Fison’s Ltd.) and Mr. G.Taylor (Messrs. Bernard Dyer and Partners Ltd.) for samples of basic slag, Mr. A. E. Heron for several helpful dixussions and Mr. H. N. Redman for most of the experimental work. APPENDIX BIAS AND ACCURACY OF THE METHOD I N FERTILISER ANALYSIS Thanks to the courtesy of Scottish Agricultural Industries Ltd. (per Mr. R. Donald) and Fisons Ltd. (Mr. E. W. Schwehr), the author is able to present the results of a collaborative trial of the above method. The details were cominunicated to these firms and collaborative analyses of a number of samples were made. Four samples of basic slag, three compound fertilisers and a sample of Morocco rock were very carefully prepared so as to be as homo- geneous as possible, divided into a number of aliquots and submitted to fifteen laboratories belonging to the three firms.Each laboratory was requested to analyse the samples for total and soluble phosphoric anhydride (water-solu ble for the fertilisers and citric acid soluble for the basic slags). Two independent determinations (not parallel determinations) were to be carried out by each laboratory on every sample both by the new and by the statutory methods, and every result was to be recorded. For the citric acid soluble phosphoric anhydride in the basic slag, however, aliquots from one citric acid extract were analysed by both methods, the whole being repeated for the second determination. A few results appeared to be inaccurate owing to “mistakes” rather than “errors” in the statistical sense of the word. In the ordinary day-to-day work of a laboratory these would have been checked, but as the laboratories had to return all results, this was not possible. If the results were “mistakes,” they have not been included in the calculationsSept., 19541 PRESENCE OF SOLUBLE SILICATES 545 of bias or standard deviation.The criterion adopted is based on the standard deviation. The mean and the standard deviation were calculated, the dubious result being excluded. If then the dubious result deviated from the mean of the remainder by four or more times the standard deviation, it was called a “mistake” and rejected. Some of the laboratories were not able to complete their tally of determinations; others, however, supplied more than two results on some of the samples. Tables IV and V show the bias and precision of the method.Samples 1, 2 and 3 are “compound” fertilisers of different types, 4 is representative Morocco rock and 5 to 8 are basic slag. The extraction procedure for “soluble P,O,” is that of the Statutory Regulations to the Fertilisers and Feeding Stuffs Act, with water as solvent for the fertilisers and 2 per cent. citric acid for the basic slags. TABLE IV AVERAGES AND BIAS OF METHOD Soluble P a 0 5 Total P a 0 6 A A < \ r \ Sample number 1 2 3 4 5 6 7 8 Number of par tici- pating labora- tories 15 15 14 15 14 12 14 13 Total number of results 101 92 50 54 97 81 93 86 Average statutory method 8.319 7.278 11.955 by 7.93 17.57 15-26 15.37 Average citric - molybdic acid method 8.296 7.267 11.914 P4: 7-88 17-49 15.09 15-27 Bias - 0.023 -0~011 - 0.04 1 - 0.05 - 0.08 -0.17 - 0.10 Average statutory method 8.958 8.995 by 33.412 9.40 19.78 16.38 16.44 Average citric - molybdic acid method 8.902 9.001 by 33.476 9.60 19.71 16.26 16-42 Bias + 0.006 +0*064 +o-200 - 0.070 -0~110 - 0.02 - 0.056 Average bias .... .. .. .. - 0.068 + 0.002 (The Morocco rock was analysed by the basic slag procedure, not the fertiliser procedure.) TABLE V PRECISION OF METHODS EXPRESSED AS STANDARD DEVIATIONS Sample number 1 2 3 4 5 6 7 8 Soluble P,O, - Standard deviation, Standard citric - deviation, molybdic statutory acid method method 0.085 0.072 0.098 0.073 0.169 0.111 0.237 0.123 0-227 0.193 0.212 0.190 0.227 0.219 Total P,O, - Standard deviation, Standard citric - deviation, molybdic statutory acid method method 0:099 0.075 0.164 0.08 1 0.228 0.250 0.289 0.177 0.320 0.178 0.261 0.169 0.276 0.144 The tables demonstrate that the bias of the new method (as compared with the old) is insignificant and that the reproducibility of the new method is better; this is more marked in the analyses of basic slag.REFEREKCES 1. Wilson, H. N., Analyst, 1951, 76, 65. 2. - , Ibid., 1949, 74, 243. 3. Stross, W., Ibid., 1944, 69, 44. 4. Anon., “American Public Health Association Methods for the Examination of Waters,” Ninth Edition, 1946, p. 44. 5. Malaprade, L., Ann. Chim., 1929, 11, 205. 6. - , Ibid., 1929, 11, 217. 7. Strickland, J. D. H., J . Amer. Chem. SOL, 1952, 74, 862 and 868.546 WILSON [Vol. 79 8. 9. 10. Muir, J. W., Analyst, 1952, 77, 313. Zimmermann, M., Angew. Chem., 1950, 62, 291. Malaprade, L., Ann.Chim., 1929, 11, 211 and 218. IMPERIAL CHEMICAL INDUSTRIES LIMITED RESEARCH DEPARTMENT UILLINGHAM, Co. DURHAM February &h, 1954 DISCUSSION MR. J. KING said that he had appreciated having a preview of the paper, so that it had been possible to compare Mr. Wilson’s method with the Official Method of the Fertilisers and Feeding Stuffs Act, 1926, jn the Government Chemist’s Department. The results by the two methods for phosphate in a standard potassium dihydrogen phosphate solution, a basic slag, it rock phosphate and a compound fertiliser agreed very closely, Mr. Wilson’s method being considerably simpler and more rapid. He noted that the mixed indicator phenolphthalein - thymol blue (the phenol violet of the British Pharmacopoeia, 1948, introduced by Kolthofi in 1927) was still preferred by Mr.Wilson t o the single indicator, in spite of its abandonment by the British Pharmacopoeia, 1953. He stated that they had used the method a t Billingham for more than 12 months for routine control analysis and regarded i t as a first-class volumetric method. MR. KING replied to Mr. Broomfield by saying that the mixed indicator was of assistance in memorising the changes of colour during the titration and in warning the operator when the neutral point was about to be reached. Under some conditions, however, notably with colourless solutions, the single indicator gave greater colour differences, and the white background of a glazed tile gave a constant reference point, which could only be obtained from reference solutions when a mixed indicator was used.He asked whether consideration had been given to the application of “The Standardisation of Volu- metric Solutions,” by the Analytical Chemists’ Committee of Imperial Chemical Industries Limited ( A naZyst, 1950, 75, 577), to the analysis of phosphates, as pure silver was readily available and much fundamental work had been done with silver by Richards, Honigschmidt and others. DR. J. R. NICHOLLS suggested that the results given flattered the official method of the Fertilisers and Feeding Stuffs Regulations. The method here described had been found to determine the whole of the phosphate without interference from silica; in the official method for basic slag there was always a discarded insoluble residue, which could be shown to contain phosphate, and the final precipitate could be shown to contain silica.As the results by the two methods were in agreement, it seemed that the two errors of the official method, by a fortunate chance, cancelled each other. DR. R. F. MILTON suggested that the references to cc and ,9-forms of silica having different rates of formation of silicomolybdate might be unproved. In his work on the colorimetric determination of silica, the silicomolybdate would form only in weakly acid solutions, and then only completely on boiling. In stronger acid solutions, e.g., N sulphuric acid, no silicomolybdate was produced, even when the solution was set aside for some time (AnaZyst, 1951, 76, 431). MR. WILSON, in a written reply to Mr. King, said that extensive laboratory experience with the mixed indicator phenolphthalein - thymol blue had shown that assistants were far less likely to overshoot the end-point than they were when the phenolphthalein was omitted. The end-point occurred a t the same pH value ; the phenolphthalein contributed to convenience. Replying to Mr. King’s question about silver phosphate, he said that it was by no means easy to prepare pure silver phosphate in quantitative yield, even from pure phosphoric acid. I t was not very insoluble in water (6.5 mg per litre a t 19-5” C), and readily soluble in dilute acetic acid or dilute ammonium hydroxide. Further, if other ions-sulphate or acetate, for example-with moderately soluble silver salts were present as well as phosphate, the silver phosphate was invariably contaminated with silver sulphate or acetate, even to the extent of several per cent., and there seemed to be no way of avoiding this contamination. In reply to Dr. Nicholls, he agreed that the good results by the statutory methods were due to com- pensation of errors. Besides the silica to which Dr. Nicholls referred, the precipitates often contained iron and lime ; the magnesium pyrophosphate obtained after dissolution of the ammonium phosphomolybdate in dilute ammonium hydroxide could contain an excess of magnesium. He had often been amazed a t the skill with which the conditions in the statutory method had been contrived so that the errors did-usually- cancel. Probably this compensation of errors, seldom quite perfect, was a cause of the lower precision of the statutory method. But to-day, when so much more was known about defects in analytical processes than was known 30 or 40 years ago, the compensation (of errors ought to be avoided whenever possible. MR. W. BROOMFIELD asked Mr. King if he could offer any better alternative. This might not always be so.

ISSN:0003-2654    

DOI:10.1039/AN9547900535

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Sept., 19541 ABBOTT AND POLHILL 547 The Determination of Copper in Oils and Fats by Means of Dibenzyldithiocarbamic Acid and its Salts BY D. C. ABBOTT AND R. D. A. POLHILL A method has been developed for the rapid absorptiometric determina- tion of copper in oils and fats over the range 0.02 to 2 p.p.m. Most of the fatty matter is removed from the sample by vaporisation and the remainder is destroyed by digestion with nitric and sulphuric acids. The copper in the resulting acid solution is then determined, after dilution, by the extraction of its dibenzyldithiocarbamate with carbon tetrachloride and measurement of the optical density of this solution a t 435 m p in a suitable spectrophotometer. The method is shown to be free from interference by some other commonly occurring metals and to give a good recovery of added copper.The preparation of the dibenzylammonium, potassium and zinc salts of dibenzyldithiocarbamic acid is also described. A determination can be completed in 2 hours. THE presence of traces of copper of about 1 part per million has been shown1 to have a marked effect on the development of oxidative rancidity in oils and fats. Although no toxic significance is associated with copper at this concentration,2 there is a need for a simple method, free from interference by other commonly occurring metals, for its determination. Many methods for the determination of copper in small amounts have been described, such as the use of sodium or diethylammonium diethyldithio~arbarnate,~~~,~ rubeanic acid,6,7 the phenanthroline~~ 7 9 and biscyclohexanone oxalyldihydrazone.1° * All of these methods suffer from the disadvantage that the sample requires neutralisation after the initial acid treatment and also often requires the addition of comparatively large amounts of inorganic salts, either to prevent interference by other metals or as buffers. The need to keep the blank to as low a figure as possible indicated the desirability of using the minimum number and amounts of reagents, and the present work was undertaken to provide a method that satisfied this condition.There was an advantage in the use of salts of dibenzyldithiocarbamic acid, the copper complex of which could be extracted from acid solution, particularly if the zinc salt was used as described by Marten and Githensll and by Stone, Ettinger and Gantz.12 I t has also been that separation of copper from cobalt, nickel, iron and manganese is favoured by extraction at a low pH.The inorganic acids used may be freed from copper by distillation. (Sulphuric, nitric and perchloric acids are available commercially as “lead free” reagents and have presumably been distilled.) The use of these reagents has made possible the reduction of the blank to about 0.3 pg of copper, and the determination of copper down to a level of 0.02 p.p.m. (0-4 pg of copper). Digestion of the oils and fats was greatly facilitated by removing most of the fatty matter by vaporisa- tion before the wet combustion, and this also led to a considerable saving of time. The loss of copper during wet combustion, referred to by Wetlesen and Gran,14 was avoided by the use of the minimum amount of perchloric acid at a stage where no free hydrochloric acid could be formed by reduction.EXPERIMEKTAL COLOUR REAGESTS IXVESTIGATED- The preparation of zinc dibenzyldithiocarbamate involved the preparation of the dibenzyl- ammonium and potassium salts of the acid, and the suitability of all these salts as reagents for copper was investigated. As there appeared to be no adequate description in the literature of the preparation of these compounds, the methods of preparation developed by us are described below. Standard amounts of copper were extracted from 50ml of 5 per cent. v/v sulphuric acid with and without the addition of potentially interfering metals. For the free acid and its dibenzylammonium and zinc salts, 10ml of carbon tetrachloride solution was used for The use of solutions of the free acid was also studied.548 ABBOTT AND POLHILL: THE DETERMINATION OF [Vol. 79 extraction.The potassium salt was introduced i n aqueous solution into the acid solution containing copper, and the copper complex was extracted with 10 ml of carbon tetrachloride. For each sample, extraction was completed by shaking for 1 minute and, after filtering through a plug of cotton-wool, the optical density of the carbon tetrachloride solution was measured in 1.5-cm diameter cells with a Unicam SP350 spectrophotometer. Results showed that solutions of the free acid and the zinc and dibenzylammonium salts (in equivalent concentrations of 0.04, 0.05 and 0.075 per cent.w/v, respectively) completely extracted up to 40 pg of copper, with no interference from 200 pg of cobalt, nickel, iron or manganese. The potassium salt, used as 5ml of a 0.01 per cent. w/v aqueous solution, gave complete extraction of copper and the interference from cobalt was slight (200 pg of cobalt G 0.4 pg of copper). Nickel, iron and manganese did not interfere at this con- centration (200 pg). Solutions of the zinc and potassium salts and of the free acid retained their power to form a complex with copper for at least 24 hours. The dibenzylammonium salt appeared to be less stable in solution. PREPARATION OF COLOUR REAGENTS- The dibenzylammonium, potassium and zinc salts of dibenzyldithiocarbamic acid were prepared as described below by methods developed from those used by Losanitschl5 for the preparation of analogous compounds.0 Dibenzylammoaium dibenzyldithiocarbamate--Add 1 a 5 ml of carbon disulphide to a solution of 5.0 ml of dibenzylamine in 5 ml of light petroleum, boiling range 40" to 60" C. Stir the mixture well until the oil that separates becomes solid (5 to 10 minutes). Add 20 ml of light petroleum and grind the solid with the flattened end of a glass rod so as to reduce it to a fine powder. Separate the dibenzylammonium dibenzyldithiocarbamate by filtration, wash it with small quantities of light petroleum -until it is free from carbon disulphide, and allow it to dry on the filter. Potassium dibenzyldithiocarbamate-To each gram of dibenzylammonium dibenzyl- dithiocarbamate prepared as described above, add 0-7 ml of a 3 N solution of potassium hydroxide in 75 per cent.v/v aqueous ethanol, and warm the mixture gently until solution is complete. Allow it to cool and add an excess of ether (10 ml for each gram of dibenzyl- ammonium salt taken) to precipitate the potassium salt. Stir the resulting slurry well, separate the solid by filtration, wash it with ether and allow it to dry a t room temperature. Potassium dibenzyldithiocarbamate is thus obtained as white silky needles of the trihydrate ; the yield is 90 per cent. of the theoretical. The trihydrate possesses no sharp melting point; it loses water of crystallisation over the range 95" to 125" C (found: K, 10.9 per cent. ; calculated for C,,H,,NS,K.3H20, 10.7 per cent.). The anhydrous material, which is formed by drying the trihydrate to constant weight in vacuo over sulphuric acid, has a m.p.205" to 206" c (loss of weight on drying, 14.4 per cent.; calculated for C,H,,NS2K.3H20, 14.8 per cent.). Cornpin16 describes potassium dibenzyldithiocarbamate trihydrate as white silky needles that melt over the range 102" to 120" C. A quantity of the dibenzylammonium salt can be recovered from the preparation of the potassium salt. Wash the ethereal filtrate well with water, separate the ether layer and allow the ether to evaporate. The dibenzylamine recovered in this way can then be treated with carbon disulphide as described aborre. Zinc dibenzyldithiocarbamate-To a solution of 0.2 g of the potassium salt prepared as above, add 4 ml of a 4 per cent. W/V aqueous solution of zinc sulphate, ZnS04.7H,0.Separate the resultant curdy precipitate by filtration, wash it with water and dry it over sulphuric acid in vacuo. Zinc dibenzyldithiocarbamate is thus prepared as a white powder having a m.p. 187" to 188" C; the yield is almost quantitative. To obtain the salt in a more crystalline form, add 95 per cent. v/v aqueous ethanol to a concentrated solution of the salt in carbon tetrachloride to the stage of incipient precipitation and set the mixture aside, when the zinc salt separates as fine white needles, m.p. 187" to 188" C (found: Zn, 10.6 per cent,; calculated for C,H2,N,S4Zn, 10-7 per cent .) . Zinc dibenzyldithiocarbamate is available commercially as a greyish-white powder having a m.p. 176" to 178" C. This may be purified by the ethanol - carbon tetrachloride procedure outlined above, when it forms fine needles, m.p.187" to 188" C. Dibenzyldithiocarbamic acid-Solutions of th.e free acid in carbon tetrachloride (0-04 per cent. w/v) were prepared by extraction with a known volume of the solvent from an acidified solution of the requisite amount of the potassium salt. The yield is 5.8 g of solid of m.p. 82" to 83" C.Sept., 19541 COPPER IN OILS AND FATS 549 METHOD All measurements of optical density are made at a wavelength of 435mp with a spectrophotometer. REAGENTS- Sulphuric acid, concentrated-Distilled or “lead free.” Nitric acid, concentrated-Distilled or “lead free. ” Perchloric acid, 60 per cent. w/w-‘‘Lead free.” Sodium sulphite solution-A 5 per cent. w/v aqueous solution. Carbon tetrachloride.Zinc dibenxylditlziocarbamate-A 0.05 per cent. w/v solution in carbon tetrachloride. Dibenxylammonium dibenzyldithiocarbamate-A 0.075 per cent. w/v solution in carbon Dibenzyldithiocarbamic acid-A 0-4 per cent. w/v solution in carbon tetrachloride. Potassium dibenxyldithiocarbamate-A 0.01 per cent. w/v aqueous solution. And one of the following colour reagents- tetrachloride . DIGESTION OF SAMPLE- Introduce about 20 g of the oil or fat, accurately weighed, into a dry 200-ml silica digestion flask. (The 200-ml long-necked boiling flasks that can be purchased from The Thermal Syndicate Ltd., are suitable. Silica flasks are preferred, as they stand up to the vigorous heating conditions better than would flasks‘imade of the usual resistance glasses.) Heat the flask until the oil begins to fume, allowing the vapours to be swept away by a good draught.Continue heating until only about 1 to 2 ml of the oil remain. Cool the flask, add 3 to 3.5 ml of sulphuric acid and then 2 to 3 ml of nitric acid. After the initial vigorous reaction has subsided, complete the digestion by heating the mixture, with the addition of further small amounts of nitric acid, until the solution is colourless when cold. (The addition of a few drops of perchloric acid is often useful in the last stages of the digestion.) When the solution is cool, add 10 ml of water to it and again heat it until fumes are evolved. EXTRACTION OF COPPER- Transfer the liquid remaining in the flask to a 100-ml separating funnel, dilute it to 50 ml with water and add 1 ml of 5 per cent.w/v sodium sulphite solution to remove traces of nitrous fumes. Add 10ml of the carbon tetrachloride solution of the colour reagent (or 5 ml of aqueous solution of the potassium salt together with 10 ml of carbon tetrachloride), stopper the funnel and shake it for 1 minute. Filter the lower layer through a plug of cotton- wool introduced into the stem of the funnel, and then measure the optical density of the extract a t 435mp. Determine the blank on the reagents under the same conditions. Determine the amount of copper present by reference to a standard curve prepared as described below. The complete determination can be made in about 2 hours. PREPARATION OF A STANDARD CURVE- Prepare a standard solution of copper by dissolving 0.157 g of copper sulphate penta- hydrate in water containing 5 ml of 5 per cent.v/v sulphuric acid and dilute it to 200 ml. This solution contains 200 pg of copper per millilitre. With suitable dilutions of this stock solution, extract the copper from a solution in 50 ml of 5 per cent. v/v sulphuric acid with 10 ml of reagent as described above, and determine values of optical density. (Measurements of optical density at 435 mp by means of a Hilger Uvispek spectrophotometer show a linear relationship to the copper content of the standard.) RESULTS The method was tested for the recovery of added copper and was compared with a direct acid extraction method. An oil-soluble copper standard was prepared by dissolving 9.4 mg of copper carbonate in 5 ml of oleic acid and diluting with olive oil to give a solution containing 10 pg of copper in 89 mg of oil.Results for the recovery of copper are shown in Table I.550 ABBOTT AND POLHILL [Vol. 79 TABLE I DETERMINATION OF KKOWK AMOUNTS OF COPPER IS OILS Copper Copper Copper Copper CLg PP P6‘ Pg Olive A . . .. .. .. .. 1.1 10-6 12.0 10.9 1.2 10-7 11.8 10.6 Olive B . . .. .. .. .. 0.3 Sunflower .. .. .. I . 0.3 10.4 10-5 10.2 Oil used in oil, added, found, recovered, W i t h aqueous potassiunr dibenzyldithiocarbamate as reagent f o r copper- By the direct acid exfvartion method- Olive A . . .. .. .. .. 1.0 10.9 11.7 10.7 Olive B . . .. .. .. .. 0-36 Sunflower .. .. .. .. 0-4 10.3 10.4 10.0 COSCLUSIONS- Comparisons have been made between the method described in this paper and one involving direct extraction by vigorous shaking with a mixture of hydrochloric acid and dilute nitric acid at 100” C.Extraction by diluted nitric acid, as is used in the determination of copper in butter,17 has been found to result in explosive ebullition with some oils.lS The recovery of copper added as the oleate to the two oils was complete, either by direct acid extraction or by the method described. For critical work, however, wet combustion is preferable as there is the possibility that copper may be present in a form that makes direct extraction by acid difficult or incomplete. We do not wish to express any preference for any of the four copper reagents used in this work, as the results were identical whichever complexing agent was used. The choice of reagent may be governed by its availability. The authors wish to thank the Government Chemist for permission to publish this paper.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. REFERENCES King, A. E., Roschen, H. L., and Irwin, W. H , Oil avd Soap, 1933, 10, 204. Ministry of Food Food Standards Committee’s Report on Copper, Analyst, 1931, 76, 554. Callan, T., and Henderson, J . A. R., Ibid., 1929, 54, 650. Sylvester, N. D., and Lampitt, L. H., Ibid., 1935, 60, 377. Ovenston, T. C. J., and Parker, C. A., Anal. Chim. Acta, 1950, 4, 136. Allport, N. L., and Skrimshire, G. H., Phavrpz. J . , 1932, 129, 248. West, P. W., and Compere, M., Anal. Chew., 1949, 21, 628. Moss, M. L., and Mellon, DL G., Ibid., 1942, 14., 931. McCurdy, W. H., and Smith, G. F., Ibid., 1952, 24, 371. Riddett, K. J., “Organic Reagents for Metals, Monograph KO. 12,” Hopkin and M‘illiams Ltd., Martens, R. I., and Githens, R. E., Axal. Chew., 1952, 24, 991. Stone, I., Ettinger, R., and Gantz, C., Ibid., 1’953, 25, 893. Irving, H. M., and Williams, R. J . P., Analyst, 1952, 77, 813. Wetlesen, C.-U., and Gran, G., Svensk Papperstidning, 1952, 55, 212. Losanitsch, S. M., Ber., 1891, 24, 3021; 1907, 40, 2970. Compin, L., Bull. SOC. Chivn. France, 1920, 27, 468. British Standard 769 : 1952, pp. 19-21. Williams, P. IS., Private communication. London, 1953. DEPARTMENT OF THE GOVERNMENT CHEMIST GOVERNMENT LABORATORY CLEMENT’S INN PASSAGE STRAND, LONDON, W.C.2 April Zzd, 1934

ISSN:0003-2654    

DOI:10.1039/AN9547900547

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Sept., 19541 YOUNG, BERRIMAN AND SPREXDBOROUGH 551 The Spectrographic Analysis of Brass and Other Materials by the Porous-cup Method BY L. G. YOUNG, J. M. BERRIMAN AND B. E. J. SPREADBOROUGH A method of spectrographic analysis making use of spark excitation of a solution as the source has been applied to the determination of major constituents of alloys. Details are given of the apparatus and method used for the determination of zinc in brass. Applications of the technique to analysis of some other alloys are described. An account is given of the successful use of added gold as an internal standard in the quantitative analysis of soluble residues, corrosion products and precipitates. THE graphite-cup - spark technique introduced by Feldman1v2 five years ago has proved a notable advance in the field of emission spectrochemical analysis in that it provides a ready means of analysing solutions with the spectrograph.In this method a small quantity of the solution to be examined is introduced by pipette into a narrow graphite cup whose porous base provides the upper electrode. The electric discharge passes between a lower electrode of solid graphite and the base of the cup, so that solution seeping through the porous graphite is fed continuously into the discharge. The porous-cup technique is capable of much wider application than the long-established mist - flame method of Lundegardh, but it is less sensitive than the copper - spark method of Fred, Nachtrieb and tom kin^.^ However, the comparatively low sensitivity is amply compensated by good reproducibility, accuracy and reliability, which commend the technique for the determination of metallic elements present in alloys in major amounts. From a spectrochemical point of view the provision of standard samples is greatly simplified by the use of a method based on solutions.Quickly available, easily prepared, graded, synthetic solutions overcome the difficulty of casting homogeneous standard alloys and free the spectroscopist from reliance on values determined by exacting and often tedious chemical analysis. Outstanding advantages of this method are that the weight of sample introduced into the source is known and is under precise control and that the sample is fed slowly into the spark over an extended period of time. Normally 0.10ml of solution containing 1.0 mg of sample is fed into the spark through the porous base of the cup over a period of approxi- mately 50 seconds, so that (assuming a constant flow of liquid) 0.2 pg of sample is excited by each of the 5000 bursts of energy occuring during this time.This paper presents details of the analytical procedure that has been adopted for use in this laboratory and also gives particulars of some further applications to the analysis of various materials. As the initial investigations were made on copper - zinc solutions, the determination of zinc in brass has been selected as a typical example. EXPERIMENTAL APPARATUS- Electrodes-Short lengths (about $ inch) of &inch diameter H.S. graphite rod (J.M.4B), supplied by Johnson, Matthey & Co. Ltd., are faced, bored and parted off in a small lathe to the dimensions given in Fig.1. The gauge shown in Fig. 2 is convenient for inspecting the thickness of the graphite at the bottom of the cup. If rather more than 1.0 mm thickness is left during the machining operation, the thickness can be made correct by scraping the graphite with a sharp knife while the cup is in the gauge, so as to produce a flat base free from contamination. A short time before use the centre of the base of each cup is made porous by placing the cup in a spare De Gramont stand with an 80" snub-pointed graphite lower electrode and passing a d.c. arc a t 3-0 amp. for 10 seconds, maintaining a gap of 2.0 mm. By means of a suitable electrode sharpener: lengths of a similar quality graphite are made with a snub-pointed tip for use as lower electrodes.Precisely dimensioned electrodes of uniform porosity are regarded as essential.552 YOUNG, BERRIMAN AND SPREADBOROUGH : THE SPECTROGRAPHIC [Vol. 79 Pipette, 0.10 ml-A pipette made from 2.0-nim bore glass tubing with the lower end drawn down to an external diameter of 1 mm for the last 20 mm and calibrated to deliver 0.10m1, is ideal for placing the solution in the graphite cup. Electrode holder-A "Barfit" electrode holder to replace the De Gramont stand has been designed for porous-cup work and has since been found to have other applications. This electrode holder is illustrated in Fig. 3, whilst dimensions and constructional details are shown in Fig. 4. t$!q . dia.. , 5 Graphite cup upper electrode electrodes adjusted ..I 1.0 mm to 2.0 mm by feeler gauge Graphite 80" snub- point lower electrode Fig.1. Details of electrodes 2: inches 1 -- (I$inchesd Gap t o facilitate 0.255'i nch diameter 1.0 mm 4 B.A. grub screw Fig. 2. Gauge for checking and adjusting graphite cup electrodes The steel foot fits into a standard Barfit clamp and carries a vertical pillar of insulating material. Electrode holders made from Q-inch square brass rod are arranged at convenient heights so that the upper one takes a graphite cup and the lower one an 80" snub-pointed graphite counter-electrode, The upper electrode holder also carries a swivelling 2-0-mm gauge of silver steel hardened and tempered to approximately 500 Brinell. The device is fixed on the spectrograph accessory bar a t the required working distance from the slit.The metal foot is raised slightly and twisted in the Barfit clamp until the electrodes are correctly disposed about the optical axis of the spectrograph, and then it is very firmly clamped. The alignment gauge,* shown to the right in Fig. 3, facilitates setting up. This electrode holder is rigid and robust, and once it has been correctly set it can be removed and replaced on the accessory bar or transported to another spectrograph without loss of alignment. METHOD FOR THE DETERMINATION OF ZINC IN BRASS SOLUTION OF SAMPLE- Dissolve 1.OOg of the brass sample (preferably drillings) in a mixture of 10ml of con- When the sample has centrated hydrochloric acid and 10ml of concentrated nitric acid. dissolved, heat the solution just to boiling, cool it and dilute it with water to 100 ml.SPECTROGRAPHIC STANDARDS- Make up standard solutions from H.S. pure metals to simulate the solution of 1.Og of copper - zinc alloy in 10 ml of hydrochloric acid and 10 ml of nitric acid diluted to 100.0 ml with water. Suitable proportions are- Copper, per cent. . . .. . . 75 68 62 55 Zinc, per cent. . . .. .. . . 25 32 38 45 Store the solutions in 4-02 glass-stoppered bottles.Fig. 3. A Barfit electrode holderSept., 19541 ANALYSIS OF BRASS AND OTHER MATERIALS 553 RECORDING SPECTROGRAMS- holder in the optical axis of the medium spectrograph, 38 cm from the slit. and photographic details are given in Table I. Set up the graphite cup (upper) and solid graphite (lower) electrodes in the electrode Optical, electrical SPECTROGRAPHIC Spectrograph .. .. .. .. Slit length . . .. .. .. Spark gap . . .. .. .. Applied voltage . . .. . . Added capacity . . .. .. Added inductance . . .. .. Upper electrode . . .. .. Slit width , . .. .. .. Distance from electrodes to slit (No quartz condenser used.) Lower electrode . . .. .. Photographic plate . . .. .. Exposure . . .. .. .. Metol (Kodak) . . .. .. Hydroquinone (Kodak) .. Developer formula- Sodium sulphite, anhydrous . . Sodium carbonate, anhydrous Potassium bromide . . .. Distilled water to . . .. Hypo (sodium thiosulphate) . . Sodium sulphite. crystals . . Acetic acid, glacial . . .. Potassium alum.. .. .. Distilled water to . . .. Micro-photometer . . .. .. Fixing bath- Lamp .. .. .. .. Slit width .. .... Slit length . . .. .. Cylindrical condenser not used TABLE I AXD PHOTOGRAPHIC DATA .. .. .. .. .. .. .. * . i5,OOO peak 0.005 pF nil .. . . .. .. .. .. .. .. .. .. .. * . . . . . * . .. .. .. .. .. .. .. .. .. * . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. } .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Hilger Medium 0.010 mm 2.5 mm 38 cm 2.0 mm Hilger standard spark unit H.S.* Graphite cup containing 0.10 rnl of solution being analysed H.S.* Graphite 4 inch diameter, shaped to 80" cone with rounded point Ilford Ordinary (Ilford Long Range Spectrum) 60 seconds 5.00 g 65-00 g 45.00 g 1-50 g 5000 ml 11.0 g 3650 g 240 g 100 ml 240 g 16,000 ml Hilger old model with Cam bridge 700-ohm d'Arsenva1 galvanometer 24 watts 0-25 mm (fixed) 2.0 cm * N.C.C.Special High Purity Graphite is used when necessary. With a pipette, place 0.10 ml of the brass solution into the cup. To avoid air pockets the pipette should initially touch the bottom of the cup and it should then be slowly with- drawn as the solution flows into the cup. After an interval of 10 seconds to allow the solution to begin seeping through the base of the cup, switch on the spark and make an exposure of 60 seconds duration. After racking down the plate, set up fresh electrodes and repeat the procedure until spectrograms have been recorded for each of the standard and sample solutions. It will be found that all the solution will seep through the porous base of the cup in 45 to 50 seconds. The solution is used up when there is a change in both the colour and the note of the spark.To ensure similarity of background density, all exposures are carried on to the full 60 seconds. PROCESSING THE SPECTROGRAM PLATE- Develop the spectrogram plate by the rocking-dish method for 90 seconds at 70" F, 100 ml of developer being used. Rinse the plate in running water, fix it in 100ml of hardening acid fixing bath for 5 minutes with occasional rocking, and then wash it in briskly running564 YOUNG, BERRIMAN AND SPREADBOROUGH : THE SPECTROGRAPHIC [VOl. 79 water for 5 minutes. it with a damp, well wrung out, chamois leather and dry it in a current of warm air. Remove the excess of water from both sides of the plate by wiping EVALUATION OF SPECTROGRAMS- With micro-photometer set to the values in Table I and a clear emulsion deflection of 50.0 cm, read the deflections of the spectral lines Zn 2502.001 and Cu 2644.802 and calculate Gauge, t inch wide x 2 mm thick, silver steel, t o swivel inch diameter L dia.steel 2 B.A. Ii-inch dia. Tufnol Fig. 4. Constructional details of the Barfit electrode holder the difference of density (log ratio Zn 2502 to Cn 2544) on each spectrogram. Prepare an analytical curve from the log ratios determined frorn the standard spectrograms in conjunction with their known zinc contents. From this curve read the percentage of zinc in the sample corresponding to the log ratio found from its spectrogram. Spectrograms from twenty samples together with four standards can be recorded on one plate. A demonstration analytical curve is reproduced in Fig. 5. The standard solutions recorded in duplicate to produce the spectrogram plate from which this curve was drawn included the four already mentioned together with six additional solutions made up to simulate brasses containing nominal small amounts of tin, lead, iron, aluminium, manganese and nickel to a total maximum of 3.0 per cent.It will be observed that the greatest deviation of any point from the mean curve is 1 per cent. of zinc.Sept., 19541 ANALYSIS OF BRASS AND OTHER MATERIALS 555 REPRODUCIBILITY OF RESULTS- The reproducibility of the method was tested by recording 32 spectrograms of an unknown brass solution on four plates. Standard brass solution spectrograms were recorded on each plate and percentages of zinc were read from their respective analytical curves, with the results shown in Table 11.The standard deviation expressed as a percentage of the zinc content approaches the usually recognised limit of precision attainable by spectro- Zinc, per cent. Fig. 6. Typical curve relating amount of zinc to the log ratio of Zn 2502.001 to Cu 2544.802 graphic methods of analysis. A chemical analysis of this sample showed that it contained: Zn, 35.59 per cent.; Pb, 0.28 per cent.; Sn, 1.40 per cent.; the remainder being Cu. As a matter of interest in connection with possible calibration drift, the log ratios of the line pair Zn 2502 - Cu 2544 for the standards included on the four plates used in the reproduci- bility test are shown in Table 111. Comparison with the analytical curve, Fig. 5, determined from a previous plate, shows little change of index or slope.TABLE I1 REPRODUCIBILITY OF RESULTS Plate number .. .. 225M 226M 227M 228M 35.0 36.0 36.5 35.4 35.8 37.3 37.4 35.8 35.5 37.5 35.3 35.3 35.8 35.5 36-3 36.2 36.5 35.3 36.9 35.5 36.0 36.0 35.5 36.4 36.0 36.8 35.6 36.2 37-0 37.3 35.5 35.3 .. . . 36-1 percent. .. . . . . Meanof 32 .. .. .. Highest figure .. .. Lowest figure .. .. .. Standard deviation . . .. . . . . 0.71 per cent. of zinc Standard deviation in terms of zinc content 37-5 per cent. (1.4 per cent. above mean) 35-0 per cent. (1.1 per cent. below mean) 1.97 per cent. Zinc content, per cent. . .556 YOUNG, BERRIMAN Ah'D SPREADBOROUGH THE SPECTROGRAPHIC [VOl. 79 TABLE I11 LOG RATIOS FROM THE REF'RODUCIBILITY PLATES Plate number .. .. .. 225M 226M 227M 228M Standard spectrogram for- 26-0 per cent.zinc . . . . 0.078 0.070 0.072 0-075 32.0 per cent. zinc . . . . 1.973 1.977 1.971 1.973 45.0 per cent. zinc . . , . 1.827 1-837 1.831 1.827 38.0 per cent. zinc . . . . 1.894 1.903 1.913 1.898 ACCURACY OF RESULTS- For 86 results so far checked on low-alloy brasses and copper - zinc solutions over the range 25 to 45 per cent. of zinc, the maximum err'or in single determinations has been 1.9 per cent. of zinc. The average of two independent determinations is normally reported. INDEX VALUE METHOD OF EVALUATIOP+- Other zinc - copper line pairs were investigated and their respective index values are shown in Table IV. Line pairs 1, 2 and 3, having a common zinc line, permit the index value method of evaluation.5 The results by this method are in good agreement with those determined by the standard log ratio procedure.This method of evaluation is very con- venient for the analysis of an occasional single sample, especially if standard solutions are not immediately available. TABLE IV ZINC - COPPER LINE PAIRS Line pair Index value, per cent. zinc Zn 2502.001 - Cu 2544.802 . . .. .. .. .. 30.0 Zn 2502.001 - Cu 2369.887 . . .. .. .. .. 65.5 Zn 2502.001 - Cu 2506.270 . . .. .. .. .. 21-5 Zn 3075-901 - Cu 3063.415 . . .. .. . . .. 37.0 Zn 2557-958 - Cu 2544.802 . . .. .. .. .. 17.0 Zn 2099.86 - Cu 2117.300.. .. .. .. .. 25.0 The above method is Suitable for low-alloy brasses containing up to 2 per cent. of elements other than copper and zinc. For the determination of zinc in high-alloy brasses, synthetic standard solutions should be made up to contain other elements in amounts appropriate to the samples being evaluated. This empirical method corrects for both the fall in copper content and possible line interference by other elements.FURTHER APPLICATIONS OF THE POROUS-CUP TECHXIQUE An extension of the range of zinc to be evaluated in copper - zinc solutions was required in connection with the determination of this element in electrolyte solutions from experimental work on the dezincification of manganese bronze.6 It was found that the Zn 2502 - Cu 2544 line pair could be utilised over the range 10 to 95 per cent. of Zn and 90 to 5 per cent. of Cu. Retaining the conditions described for the -solution of the sample and the preparation of spectrograms, this technique has been used to determine other single Jements, as illustrated by the following examples- (i) Small amounts (0.001 to 0.3 per cent.) of magnesium in iron making use of the line pairs-Mg 2795.55 - Fe 2788.105, i:ndex 0.014 per cent.Mg. Mg 2802.695 - Fe 2806.985, index 0.005 per cent. Mg. (ii) Aluminium (0.10 to 20 per cent.) in copper making use of the line pair A1 3092.713 - Cu 3063.415, index 0.25 per cent. Al. (iii) Aluminium (5 to 95 per cent.) in alurrlinium - magnesium alloys making use of the line pair Mg 2928-75 - A1 3082.155, index 60 per cent. Al. In this determina- tion a 0.1 per cent. solution of the sample and synthetic standard solutions were prepared.Sept., 19541 ANALYSIS OF BRASS AND OTHER MATERIALS 557 If similar standard conditions and an appropriate series of synthetic standards are used, the method is suitable for application to many other metallic alloys. With the increased dispersion provided by the large quartz spectrograph, the analysis of solutions of steels is practicable, and the following line pairs have been selected from spectrograms recorded from solutions of the B.C.S.Low-alloy Steel Series- Mn 2933.063 - Fe 2936.905, index 0.9 per cent. Mn Ni 3101.554-Fe 3100.666, 33 3.0 per cent. Ni Cr 3593.488-Fe 3594.636, 97 0.1 per cent. Cr 310 2816.154 - Fe 2813.613, 33 0.45 per cent. Mo V 3102.290 - Fe 3100.666, 33 0.29 per cent. V Cu 3273.962 - Fe 3277.346, '3 0.08 per cent. Cu ADDED GOLD ISTERNAL STAXDARD As the results were promising when the porous-cup technique was applied to metallurgical analyses, it seemed profitable to attempt the quantitative analysis of powders, scales, residues, precipitates and corrosion products by the same method.By their nature these samples are frequently very small, and usually only a qualitative analysis by globule arc is attempted. As the arc source has poor reproducibility, spectrograms of such samples recorded by means of a solution - spark technique would be expected to be more reliable quantitatively, although less sensitive to traces of elements. Difficulties in connection with the solution of an unknown sample must be accepted. Corrosion products are frequently soluble in dilute aqua regia, while scales and residues are usually only partly soluble. Some chemical preparation and probably separation is frequently inevitable, and each individual sample must be treated in accordance with its characteristics.Another difficulty arises from the fact that advance arrangements to use a major con- stituent of an unknown sample as an internal standard cannot be made. Instead, it is possible to add a definite amount of some element not present in the sample to act as an internal standard. The precise addition of an internal standard element is particularly easy when the sample being analysed is in, or may be brought into, a solution of known weight in known volume. An element to be used as an internal standard in this connection should, if possible, possess the following desirable characteristics. (a) The element should not be liable to occur, even in small quantity, in any samples that might be encountered.( b ) It should be readily obtainable in the pure state, be soluble in dilute acids and form salts that are soluble in water; it should not react in solution with other elements being determined. (c) It should give from 30 to 60 spectral lines, dispersed throughout the normal photo- graphic region, in groups nicely graded in density. (d) It should not interfere with either the emission or the spectral lines of other elements. ( e ) Excitation potentials should be average and the intensities of spectral lines should remain .reasonably constant in spite of small changes in excitation conditions. After considering the various elements available and the indexes of individual element lines given by Brode,' platinum, gold and mercury were selected as possible internal standards.Dilute solutions of these three metals were made, and their spectra were recorded on an Ilford Long Range Spectrum plate by the porous-cup technique under the conditions shown in Table I. Examination of these spectrograms indicated quite definitely that gold showed the greatest promise, although it still fell short of the ideal requirements mentioned above. Before further experimental work was proceeded with, it was thought advisable to acquire and record precise information regarding the gold spark spectrum under the given standard conditions. Sixty-eight lines for gold were recorded ; these are listed together with their measured densities in Table V. A working standard gold solution was made up from Matthey high-purity metal to contain 2.00 per cent.of gold, 10 per cent. of concentrated hydrochloric acid and 10 per cent. of nitric acid. Dilution of a sample solution with an equal volume of this gold solution would558 YOUNG, BERRIMAN AND SPREADBOROUGH : THE SPECTROGRAPHIC [VOl. 79 thus ensure a final gold content of 1.0 per cent., Le., 1.0 mg of gold in each 0.10 ml of analytical solution taken in a graphite cup for excitation. Standard synthetic solutions were made up from various elements and diluted to give a series of solutions of graded element content, the gold content remaining constant at 1.0 per GOLD LINES I I I I1 I I I1 I1 I I1 I I I I I, I1 I I I I I1 I1 I, I1 XI I, I1 I I I1 TABLE V OCCURRING I N POROUS-CUP TECHNIQUE SPECTROGRAMS UNDER RECOMMENDED CONDITIONS THE 481 1.62 4792.60 :EE} 4084-12 4065.08 4052.81 4016.05 3897-89 3633.24 3607.54 3565.93 3553.572 3320.151 3308.310 3267.08 3230.636 3204.74 31 94-71 3122-781 3029.205 2994.99 2990.28 2954.39 2932-19 2913.54 2907-06 2905.90 2893.42 2891.96 2883.45 2847.09 2833.64 2825.45 Density 0.08 0.16 0.42 0.47 0.63 0-37 0.28 0.45 0.21 0.23 b.v.0.35 0.21 0.17 b.v. 0.23 0.19 0.15 0.82 0.40 0.37 0-38 0.20 0.22 0.57 0.26 0.19 0.14 0.15 0-18 0.11 0.12 0.29 I, I1 I I I I I, I1 I I1 I I I I I, I1 I1 I1 I1 I1 I1 I1 I1 cent.; a normal series consisted of 330, 100, 33, 10, 3.3, 2822.72 2819.95 2802.19 2780.83 2748.26 2700.89 2688.71 2687-63 2675.95 2641.49 2590.04 2565.71 ;E:;:} 2510.49 2503.32 2427.95 2387.75 2376.24 2364.56 2352.65 2340.19 2322.28 2315.85 23 14-64 2309-43 2304-81 2291.52 2283.32 2277-64 2242.71 2229.03 2201.35 2110.80 Density 0.19 0-26 0-45 0.11 0.77 0.1'7 0.17 0.11 0.98 0.40 0.10 0.09 0.08 0.11 0.25 1.01 0.18 0.07 0.10 0.26 0.06 0.07 0.07 0.08 0-05 0.15 0.09 0.26 0-04 0-16 0-11 0.13 0.07 1.0, 0.33 and 0.10pg of element per Oil0 ml of solution.The dilution was greater for elements more sensitive tothe spectro- graph. Spectra were recorded for these synthetic standard solutions on an Ilford Long Range Spectrum plate under the conditions shown in Table I. Gold - element line pairs were selected and measured with the micro-photometer, and families of analytical curves were drawn. Some referring to iron are shown in Fig. 6, and the index values noted are in Table VI. Families of curves and index values have been determined for Al, Ba, Ca, Cr, Co, Cu, Mg, Mn, Ni, Pb, Si, Sn, Sr, V and Zn.Standards were made up and conditions were chosen on the assumption that in making an analysis the final solution used would contain 2 per cent. of sample, 10 per cent. of con- centrated hydrochloric acid and 10 per cent. of nitric acid. A small accurately measured volume (0.50 ml) of this solution could be mixed with an equal volume of 2.0 per cent. gold solution so that the final 0.10-ml aliquot excited .to produce the spectrogram would contain 1.0 mg of sample and 1.0 mg of gold. OPERATION OF THE GOLD INTERNAL-STANDARD METHOD Solutions from mixtures of a few elements with not too complex spectra can be analysed Element amounts or percentages quantitatively under the conditions described above.Sept., 19541 ANALYSIS OF BRASS AND OTHER MATERIALS 559 read by interpolation between previously standardised index values are conveniently deter- mined with the Judd Lewis comparator or, if a spectrum map of line pairs has been prepared, the spectrum projector. For the highest accuracy, micro-photometer interpolation from a repeat plate containing spectrograms from the sample and appropriate synthetic standards Iron, fig Fig.6. Curves for iron - gold solution: curve A, Fe 2404-882 - Au 2387.75; curve B, Fe 2714.412 - Au 2700.89; curve C, Fe 2755.737 - Au 2748.26; curve D, Fe 2947.658 - Au 3029.205; curve E, Fe 2947.658 - Au 2913.54; and curve F, Fe 2936.905 - Au 2913-54 is recommended. When the forms of the analytical curves of the line pairs are known together with the approximate amount of the element being determined, then spectrograms from only two synthetic standard solutions are required.Except for relatively simple mixtures of elements, the spectrogram at medium dispersion is too crowded for convenient evaluation with the Hilger micro-photometer. Use of the Zeiss micro-densitometer considerably increases the convenience of evaluation of a complex spectrogram. If sufficient time is available, then really complex spectra can be made to yield TABLE VI IRON - GOLD LINE PAIRS Line pair Fe 2382.04 - Au 2378.75 . . Fe 2404.882 - Au 2387-75 . . Fe 2599.396 - Au 2641.49 . . Fe 2611.872 - Au 2641-49 . . Fe 2628.292 - Au 2641-49 . . Fe 2714.412-Au 2700.89 .. Fe 2755.737 - Au 2748.26 . . Fe 2755.737-Au 2780.83 ..Fe 2936.905 - Au 2913.54 . . Fe 2945-055-Au 2913.54 .. Fe 2947-658-Au 2913.54 .. Fe 2947.658-Au 3029.205 .. Fe 3020.64 -Au 3029-205 .. .. .. .. .. .. .. . . . . . . .. .. .. .. Index value Amount of Element in I A -I element, 1-O-mg sample, PCg Yo .. 0.78 0-078 .. 1-60 0.16 .. 2-40 0.24 .. 6-60 0-66 .. 21-0 2-10 .. 11.2 1.12 .. 35-0 3-50 .. 0-24 0-024 . . >400-0 > 40.0 . . 350.0 35-0 . . 245.0 24.5 . . 120.0 12.0 .. 33.0 3.30560 YOUNG, BERRIMAN AND SPREADBOROUGH THE SPECTROGRAPHIC [VOl. 79 accurate results. The effort is justified when an analysis cannot be made in any other way. Dilution of the sample content of an analytical solution whilst the gold is kept constant a t 1.0 per cent. can be used to simplify a crowded spectrogram, but this procedure would not be expected to reduce line interference.Ideally the porous-cup technique with gold internal standard should be used in conjunction with some degree of chemical separation, following which the solutions can in turn be applied to the spectrograph for element deter- mination. When it is necessary to record the whole of the photograpjhic spectrum with this instrument, three The large quartz spectrograph has been used for this purpose. I .2 I .o 08 x c U v) .- f3 0 6 04 02 0 I I 0 Indium, Ag Fig. 7. Calibration curve relating amount of indium to density of line In 3256.090 consecutive exposures can be given on different wavebands , successive quantities of sample solution being used in the same porous cup. Unfortunately the spark source between graphite electrodes gives rise to considerable photographic background owing to light reflected at the first surface of the Littrow collimator.Experimental work is continuing with a prototype model Cornu spectrograph of increased dispersion. RECENT DEVELOPMENT- Provided care is taken to see that identical volumes of solutions are placed in the porous cups with the pipette, and sparking is continued until the solutions are all consumed, it has been demonstrated that analytical curves from element line density against element weight can be drawn. Such a curve is shown in Fig. 7. It will be appreciated that the deviation of individual points from a smooth curve is primarily governed by the precision with which identical volumes of solutions are introduced into the porous cups and excited in the source. This is, in effect, the total-energy method suggested by Slavin.8 A spectrographic technique freed from the restrictions imposed by the conventional internal standard method of evaluation would simplify the analytical procedure and reduce the possibility of line interference. A pipette has been produced that will deliver 0.10 ml with a precision of &-O.OOl ml. This allows element determination when standard spectrograms are included on the same plate. One step in this direction has been made. This paper is published by permission of the Admiralty.Sept., 19541 ANALYSIS OF BRASS AND OTHER MATERIALS REFERENCES 1. 2. - , Ibid., 1949, 21, 1211. 3. 4. 5. 6. 7. 8. Feldman, C., Anal. Chem., 1949, 21, 1041. Fred, M., Nachtrieb, N. H., and Tombs, F. S., J , 09t. SOC. Amer., 1947, 37, 279. Young, L. G., Engineer, 1944, 128, 116. -- , Ibid., 1949, 187, 589. McClune, Helen M. C., Thesis, 1950, Queen’s University, Belfast. Brode, W. R., “Chemical Spectroscopy,’’ John Wiley & Sons, New York, 1943. Slavin, M., Ind. Eng. Chem., 1938, 10, 407. 561 ADMIRALTY MATERIALS LABORATORY HOLTON HEATH POOLE, DORSET Mavch loth, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900551

出版商:RSC    

年代:1954

数据来源: RSC