Document DMarQBKkq111QRRwQ67dXQ3QQ

United Kingdom Atomic Energy Authority RESEARCH GROUP Memorandum !i- : r' \ '' .' ' '' ' > *'V'- A SIMPLE APPARATUS v FOR THE GAS CHROMATOGRAPHY ' OF POLYPHENYLS ; /.; ... " . . -9 - .. " ': v i : : . .' . ' > ft. W, WILKINSON J. A. WINTER "V-v V :.. . - . .y^ . 7/ '.< -jcr V 1; Chemistry Division, Atomic Energy Research Establishment, ... ` ' ' , ^ ' ' ' 'iff- 1 Harwell, Berkshire. .1961 Available from H.M. Stationery Office PRICE 2s. d. NET id'afiu ttULtawM*iitf6il<JiV,'a 30067 , TOWOLDMONOQ56671 , (c)- UNITED KINGDOM ATOMIC ENERGY AUTHORITY - 1961 Enquiries about copyright and reproduction should be addressed to the Scientific Administration Office, Atomic Energy Research Establishment, Harwell, Didcot, Berkshire, England. . '; , ' ; ' ` . 'i . x u.o.c. : ' 5<7.1:$,5M.25 , S4J.544.25:S47.M1 0067815 TOWOLDMONOQ56672 UNCLASSIFIED (Approved for Sale) AEHE - M.820 A SIMPLE APPARATUS FOR THE CAS CHROMATOGRAPHY OF PQLYPHENYLS l>y R. W. Wilkinson and J. A. Winter ABSTRACT A simple and oheap gas chromatography apparatus for the analysis of polyphenyls has been developed. It has been designed to deal with both solid and semi-liquid samples. The apparatus makes use of model engine glow plugs for the detectors and a column of alumina with no liquid phase. The normal running temperature is U25C. Chemistry Division, U.K.A.E.A. Research Croup, Atomio Energy Researoh Establishment, Harwell. February, 1%1 HL 61/1004 (SC 2) 00 fc?8U> TOWOLDMONOQ56673 CONTENTS Page No. 1. Introduction 1 2. Fumaoe 1 3. Deteotor Unit and Column X 4. Deteotor Circuit 2 5. Sample Injeotion 2 6. Qualitative uee 2 7. Quantitative uee 3 6. Acknowledgements 3 9. References 3 10. Results 4 Tables I Analysis of Santowa*-R (Batoh l)By Oas Chromatography 4 II Analysis of various SynthetioPolyphenyl mixtures containing High Boilingfractions 5 , 0067817 TOWOLDMONOQ56674 1. Introduction A simple laboratory gas ohromatography apparatus suitable far routine determinations of polyphenyls, and especially for the analysis of Santowaxes has been developed in the Radiation Chemistry Group. The apparatus is constructed from readily available oexponents, is oheap to assemble, robust, and simple to operate. The sample injection system has been designed to deal with solid and semi liquid samples, since these are the two forms in which polyphenyl mixtures are most commonly available. Irradiated samples especially are prone to be semi liquid. Basioally the apparatus consists of a laboratory furnace on end maintained, at about 425C, a Felton 0) glow-plug detector, and a stainless steel oolumn paoked with ordinary ohromatographio alumina using helium as carrier gas. 2, Furnace The fumaoe consists of a 3" I.D. stainless steel tube wound with insulated 20 SWG nichrome wire and enclosed in the usual lagged Sindanyo box; steady temperature conditions are obtainable by means of a 2 kw. Variao run directly from the mains. Our apparatus runs at a very steady ltZ5C at 95 volts, the only adjustment to the Variao being very slight to cope with seasonable variations in the ambient temperature. The temperature variation over the length of the columndetector unit, which occupies 15" of the furnace length, is 1093, and the temperature quoted is that of the central portion of the column. The helium in let tube is preheated with a standard heating tape run off D.C. mains voltage. 3. Detector Unit and Column The detector is basioally the same as that described by Felton ^ and employs hot wires in the usual manner, the novelty being the use of model aircraft glow plugs as the hot wires. It is essential to use ceramic insulated plugs (e.g. Champion V.G. -2) and not fibreglass insulated (such as K.L.G. Miniglow) as the insulation burns away in the latter. The layout of the detector block and gas connections is shown in Figs. 1 and 2. The connections to the glow plugs are by end silver soldering 20 S.W.G. silver wires on to the top terminal of the plugs, after which they must be boiled in water for an hour or so to remove all traces of flux. If any flux remains the joint will last only a short time. The plug con nections are covered in ceramic beads and are taken out through holes in the top Sindanyo oover plates. They are taken well away from the furnace before good soldered connections to the bridge circuit are made. The entire detector blook and gas inlet tubes are packed thickly round with wet asbestos pulp, and the whole top plate on the outside of the furnace is covered with mare asbestos pulp which will dry quite hard and strong. It is essential that the top plate should be well insulated thermally, especially where the detector leads are brought out from the furnace. The leads too must be well Insulated, since a really good baseline will not be obtainable if these simple precautions are not taken to eliminate variations in the resistance of the leads. -1 006/618 The deteotor block itself consists of a 2" by 1" by 3" stainless steel block cut from standard 2" by 1" bay and bored to take jf" O.D, stainless steel stubs, welded in. The glow plugs are screwed |p" by 32 t.p.i. and are fitted in their holes so as not to project into the gas stream. The most satisfactory if" unions are 'Simplifix' stainless steel fittings inside the furnace, and brass 'Simplifix' fittings outside. For the fittlngB inside the furnace, the use of anti-scuffing compound on the threads will assist dismantling whenever replacement of the column is desired, the fittings inside the furnace being used once only, as leaks occur if fittings are "second-hand". The oolumns are constructed from standard bending quality if" O.D. stainless steel tube, and are packed with B.D.H. Chromatographio alumina, the ends of the column being plugged with tufts of silioa wool. The oolumns are highly retentive, and the retention times are quite sensitive to column length. We have found 22-24" to be a good compromise between satisfactory resolution of the terphenyls and time of analysis. The column is wound with thick asbestos tape from the detector block downwards to assist even temperature distribution. 4. Detector Circuit The detector circuit is extremely simple (Fig. 3) and consists of a straight forward Wheatstone net fed by 2 volt 150 A-E accumulator, the uribalanoe voltage being taken to a 5 or 10 mV recorder, the accumulator being trickle charged during use. The Honeywell Brown 2,5 mV recorder fitted with 10, 5 and 2.5 mV ranges has been found to be very satisfactory. The detector is not highly sensitive, as one would expect, the output being of the order of 2.0 mV/mg for biphenyl, and sonwhat lower for the higher polyphenyls. However, since a sample of the order of 15-18 mg is required to give peak areas on the linear part of the calibration graph, the sensitivity is quite satisfactory for normal use. Also, with these mixtures, the composition of very small samples might not bo accurately representative of the whole mixture. 5. Sample Injection A sample of about 10 to 15 mg is pushed into the small hole at the base of the sample injector (Fig. 4) which can be weighed before and after; a thread is provided to sorew into the silver steel rod. The rod is then pushed through the silicone rubber washer (Fig. l) and the brass tap turned to the open position. The injector rod can then be Inserted to a point directly over the column. The silicone washer ensures a gas tight seal round the silver steel bar. The whole operation should be oarried out without unnecessary force as a hard push may dis lodge a portion of the sample, thereby giving an incorrect result for quantitative analysis. 6. Qualitative U3e A large range of stable hydrooarbons give identifiable peaks; up to para quater-phenyl. Retention time for a given substanoe is not a oonstant, but for a given unknown mixture spiking with known substanoes gives a first indication of the identity of unknown peaks. This method is very simple and with biphenyl and the terphenyls very rapid (see Fig. 5), but the quaterphenyls take some time. - 2- 0067819 TOWOLDMONOQ56676 about one hour for p-quaterphenyl. The peaks from the quaterphenyls are much flatter than the corresponding terphenyls, but are easily observed (see Figs. 6 and 7)* Some work has been done on mixtures of naphthalene, anthraoene and phenanthrene and these were resolved as shown in Fig. 8. Higher poly phenyls such as m-eexiphenyl have not as yet been observed as peaks but a hump in the base line occurs when a sexiphenyl is passed through the column and the base line requires some time to return to normal. The polyphenyl hydrocarbons used have given no evidence of decomposition, such as multiple peaks. The type of ohromatogram produced is shown in Fig. 5* 7. Quantitative use The peak area from a given component depends on the mass present but is also affected by the presence of other components. However, a known weight of a given mixture will always produce the same peak areas under the same conditions and the quantitative use is based on this. The peak areas are measured with an Amsler planimeter. To calibrate the apparatus a polyphenyl wax of known com position was prepared by weighing out accurately various polyphenyls, well mixing and melting very carefully. An excellent wax can be made up from 5$ by weight of Biphenyl, 10$ of o-Terphenyl, 40$ of m-Terphenyl, and 45$ of pTerphenyl. Varying known weights of this wax can be placed in the sample injeotor and passed through the column. The areas of the peaks produced for each polyphenyl can be plotted against weight in mg. and then used as calibration graphs. A note of the pressure of the helium flow and the detector volts must be taken to ensure that conditions may be reproduced, also the column temperature must be fixed at 425C. The helium flow through the column is never turned off but left flowing night and day, and this enables one to use the apparatus without unnecessary delay. When an analysis has been made on a polyphenyl mixture a check on the acouracy can be made by making UP a known mixture with the composition of the polyphenyl mixture upon which the analysis was done. This mixture is then run through the column and the peak areas measured as before. Thus if necessary a correction can be made to the original analysis. It has been observed that with mixtures of higher boilers and biphenyl/terphenyl mixtures the difference between the amount of high boilers present and those indicated increases with increase of high boilers (Fig. 9). By successive approximation it is possible to make a known mixture with same peak areas as the unknown, thus taken to have the same composition. Table I shows a typical analysis on Santo-Wax R polyphenyl wax. 8. A cknowledgements The authors are grateful to Dr. W. G. Bums for much helpful advice and to Mr. B. Morris for many practical suggestions. Thanks are also due to Dr. J. Cade for the supply of pure quaterphenyls. 9. References 1. Felton H. R. GaB Chromatography. Chapter XV. York. 1958, page 131 - 143. - 3- Academio Press New 00b?z0 TOWOLDMONOQ56677 10. Results Tifflra t Analysis of Santovax-R (Batch 1) by Gas Chromatourauhv No. of Run Weight of sample ags Biphenyl Weight 9? o-Terphenyl Weight % m-Terphenyl Weight % p-Terphenyl Weight % I4B.3. by difference 1 2 3 4 5 6 7 8 MEAN 12.6 16.1 14.0 14.0 U.3 15.0 12.0 12.4 - 1.27 1.18 1.28 1.29 1.15 1.28 1.10 1.21 1.22 10.3 9.0 8.6 10.0 9.15 8.6 8.5 9.0 9.14 56.lt- 58.2 56.0 55.0 53.0 53.0 55.0 54.2 55.10 23.0 23.0 26.0 24.2 26.5 25.3 22.6 23.8 24.30 9.03 8.62 8.12 9.15 9.90 11.82 12.8 11.79 10.24 Standard Deviation 0.07 0.66 1.76 1.47 1.71 TOWOLDMONOQ56678 TAHT.E TT Analysis of Tarlous Synthetic Polyphegyl Mixtures contain-? gjghjfrollingfractions Run Ho. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Height of saaple in ags 14.5 12.0 12.8 15.3 14,5 14.8 18.4 17.0 19.0 18.2 18.2 20.0 18.2 20.0 19.4 Biphenyl found % Biphenyl present % 1.4 1.22 1.2 1.22 1.3 1.22 3.2 3.9 1.3 1.5 1.3 1.5 0.95 1.0 0.89 1.0 1.10 1.0 2.70 2.97 2.70 2.97 2.50 2.97 1.3 1.5 1.3 1.5 1.3 1.5 t^Terphenyl found % o-Terphenyl present % 9.0 9.14 9.0 9.14 9.2 9.14 7.6 7.9 8.0 8.5 8.1 8.5 6.4 7.0 7.05 7.0 6.85 7.0 5.30 6.01 5.80 6.01 5.40 6.01 5.2 6.5 5.4 6.5 6.2 6.5 a^Terphenyl found % -Terphenyl present % 54.4 55.4 54-8 29.0 55.14 55.14 55.14 31.0 45.5 45.0 45.2 45.0 40.5 42.0 39.0 42.0 40.5 42.0 20.50 23.58 20.50 23.58 20.35 23.58 16.7 22.0 16.9 22.0 18.0 22.0 p-Terphenyl found % p-Terphenyl present 24.0 24.3 23.8 24.3 24.0 24.3 35.6 35.9 21.0 25.0 22.4 25.0 18.5 20.0 18.0 20.0 17.55 24.5 20.0 27.31 24.5 27.31 24.65 27.31 18.1 20.0 18.4 20.0 15-5 20.0 High Boilers found % difference High Boilers present* 11.2 10.2 10.5 10.2 10.7 10.2 24.6 21.3 24.2 20.0 23.0 20.0 33.65 35.06 34.0 30.0 30.0 30.0 47.0 40.13 46.5 40.13 47.1 40.13 58.7 50.0 58.0 50.0 59.0 50.0 Excess Kean % High Boilers found 0.8 3.50 4.25 6.73 8.56 * High Boiler Content made up with p-quaterphenyl. ? ? 8 /900 TOWOLDMONOQ56679 SILICONE RUBBER WASHER 6?8^3 TOWOLDMONOQ56680 TOWOLDMONOQ56681 END STOP II % SILVER STEEL ROD SAMPLE BRIDGE CIRCUIT AERE - M 820 Fig. 3 SILICA WOOL PLUG AERE - M 820 Fig. 4 SAMPLE INJECTOR 00b7?* TOWOLDMONOQ56682 AERE - M 820 Fig. 5 0067826 TOWOLDMONOQ56683 DATE. 4-9-GO. SAMPLE PILE IRRADIATED. NEW S.W. ft WEEKS. COLUMN. ALtOj ll\ AMOUNT. SSmgms. TEMP.C. 415. PRESSURE. 10 !*>. VARIAC. 99 RANGE. SOmv. DET. VOLT6.M5. SPEED. 9* hi-. Z I 0tt. w El x j att so >Z iHn hi aX O a J > ui aX . * X l A. _l________ I__________L_ S9 Sft 49 _iiii 4t 95 S9 SI AERE * M 620 Fl|. 6 iImins. 14 1 0067827 TOWOLDMONOQ56684 DATE. 47-G-GO. COLUMN. ACgO ZZ\ TEMP. C. 425. VARIAC. 92. DET. VOLTS, I-S MIXTUttlOF O.M fcNOP )AMrLc. QUATER PHENYLS. AMOUNT. I0r*$m%. PRESSURE.!*! lb! RANGE. 4* 5 my. SPEED 5"hK a K J > UZl Zo aX at 3 O o 90 *0 AERZ - M 820 Fig. 7 * THE OUATERPHENYLS INJECTED ARE THOSE LINKED IN THE O.M AND P POSITIONS. IS 0 MINS. 0067828 TOWOLDMONOQ56685 DATE. 9-12-GO. COLUMN. Aet03 2 2 TEMP. C. 425. VAR I AC. 92. DET.VOLTS.I-G. SAMPLE. AMOUNT. lOmgmft PRESSURE.IOIbs .' RANGE. 5mv. SPEED. 8V. o o o NAPHTHALENE t II 21 14 7 AERE - M 820 Fig. 8 MIN6. TOWOLDMONOQ56686 AERE - M 820 Fig. 9 0067 830 TOWOLDMONOQ56687 Available from HEfc MAJESTY'S STATIONERY OFFICE York House, Klngsway, London W.C. 2 423 Oxford Street, London W. 1 , 13a Castle Street, Edinburgh 2 109 St. Mary Street, Cardiff 39 King Street, Manchester 2 50 Fairfax Street, Bristol 1 2 Edmund Street, Birmingham 3 t ` 80 Chichester Street, Belfast or through any bookseller. Printed in England S. O. Code No. 91-3-10-61 ) TOWOLDMONOQ56688