1420 THE COLLIERY GUARDIAN. June 19, 1914. Removal of Carbon Monoxide from Air, and Methods of Estimation in Mines. By JOHN HARGER, M.Sc., Ph.D. From a paper read before the Manchester Geological and Mining Society. Part I. There are many methods of estimating carbon monoxide, but the objection to them is that in most cases they cannot be applied “ on the spot ” in mines, at places where the estimation would be of great value. The use of blood solutions for detecting carbon monoxide was the basis of Vogel’s method some 40 years ago. This change was afterwards rendered a really delicate test by the discovery that the “ absorp- tion bands ” of the oxyhaemoglobin were altered by treatment with reducing reagents, but the “ bands ” of the compound formed from carbon monoxide with blood—the carboxyhaemoglobin—persisted after such treatment. Dr. J. S. Haldane has done much work on the blood method of estimating carbon monoxide. The blood method is chiefly of physiological interest, and the investigations have given some useful informa- tion; but the writer thinks that mining men will agree that it can never be of any real use in mines, and having done its work from the practical standpoint, it must be regarded as useless for present needs. The writer is only concerned in this paper with the methods which are suitable for estimating such small quantities of carbon monoxide as are detrimental to health, from 0-01 upwards; and for this 0-01 per cent., and up to 0-4 per cent., the ordinary cuprous ammonium chroride is useless, as the volume change is too small. There remain four other known processes, namely :— (1) Reduction of gold chloride solution; (2) Reduction of palladium chloride solution ; (3) Reduction of iodine pentoxide (I205) (Nicloux* process); (4) Reduction of yellow mercuric oxide. The first 'two methods are based upon the precipita- tion of the gold or palladium, and the metal may be weighed, or for traces of carbon monoxide the solutions show changes of colour : a drop of palladium chloride (PdCl2) solution, for instance, darkens with a trace of carbon monoxide in the air. A scheme can be worked out to ascertain the percentage of carbon monoxide in the air by observing the time taken to darken to a certain standard depth of colour. This is a good test in the laboratory with daylight, but the writer has found it of little use in the mines, because dust and artificial light (both variables) rendered it very uncertain. The third method—the reduction of iodine pentoxide (I2O5)—appeared at first to be worse than hopeless, as this material has to be heated to 120 to 170 degs. Cent, for the reaction to take place rapidly with carbon monoxide—a process which is quite easy in the labora- tory, but which appeared an absolute bar for its use in mines. It was found, however, that the space above the outer gauze and inside the bonnet of a safety lamp was just what was required, the temperature there being about 170 degs. Cent, with the lamp full on and not “ smoking.” It appeared almost as if the safety lamp had been designed for this part of the test. In the iodine pentoxide test the iodine set free is a measure of the carbon monoxide— I205 + 5CO = 5CO2 + I2 Iodine pentoxide Carbon monoxide Carbon dioxide Iodine and the iodine can be so accurately estimated and the end reaction seen in such poor light that it is quite unmistakable. A mere trace of free iodine gives an intense blue colour with starch. If a dangerous quan- tity of carbon monoxide is present, warning is given within a minute by this test; for less quantities a longer time is taken to obtain the result, which is just what is wanted in a test of this kind. Another method, which is a modification of that just described, but which is better in some ways, is to make use of the fact that carbon dioxide is evolved by the reaction of carbon monoxide with iodine pentoxide— I2O5 + 5CO = 5CO24-I2. In this process the iodine is kept back by passing the gas after the iodine pentoxide tube over filings or turn- ings of metallic copper. The gas containing the carbon dioxide is then passed through a spiral tube containing a definite quantity of standard solution of caustic potash (or soda), followed by a small vessel containing a trace of baryta water coloured by phenolphthalein. This colour is very intense, and as soon as the carbon dioxide (from the carbon monoxide, if present in the air) has used up the caustic potash (or soda) in the spiral, it comes forward to the vessel containing the baryta and phenolphthalein. When this hap- pens, the violet colour disappears, and the percentage of the carbon monoxide is read off on the rod of the piston as in the starch iodine reaction already described. An automatic method for recording carbon monoxide in air or in hydrogen has been worked out, and it depends upon the fact that dilute solutions of hydriodic acid produce small primary currents, which are exactly in proportion to the strength of the hydriodic acid when brought into contact with copper and zinc in a small cell, the copper and zinc are connected through a record- ing galvanometer. The air or gas to be tested is passed through drying tubes, and then through a tube contain- ing iodine pentoxide heated in the case of carbon monoxide in air to 170 degs. Cent., and in the case of carbon monoxide in hydrogen to 130 degs. Cent. The * Gomptes Rendus, 1898, vol. cxxvi., p. 746. carbon monoxide reacts with the iodine pentoxide, giving iodine and carbon dioxide— I2O5 + 5CO = 5CO2 + I2. The hot gas containing the iodine is passed into an absorption apparatus, in which it is brought into inti- mate contact with arsenious oxide solution. The iodine is converted into hydriodic acid by the arsenious oxide, and the liquid containing it is passed in a slow con- tinuous stream through the small cell containing copper and zinc previously mentioned, and the electric cur- rents (and so the percentage of carbon monoxide) is recorded from time to time at pre-arranged intervals by the recording galvanometer. The gas from which the iodine has been removed is then passed through a gas meter (of the wet type) provided with a governor which regulates. the speed to one revolution each 10 or 15 minutes. The flow of liquid is made quite constant by fixing to the spindle of the meter drum two spirals (made of glass or metal), which take up exactly the same amount of the solution from a constant level vessel each revolution of the drum and discharge it through the absorption apparatus. The rate at which the carbon monoxide is passed depends upon several factors, and can be varied to suit requirements. Gener- ally a litre of gas per half-hour has been used in the tests, and for carbon monoxide in hydrogen this is rather quick for the size of iodine pentoxide tube used. Twice the rate can be used for twice the surface of iodine pentoxide exposed, etc. Hydrogen itself reacts slowly with iodine pentoxide, but at 170 degs. Cent, the rate at which it is attacked is about 3-J0 of that at which carbon monoxide is affected. In determining the carbon monoxide in hydrogen, this factor cannot, however, be overlooked, and the zero must be determined by using hydrogen carefully purified by sending it through a “blank” tube of iodine pentoxide. For this reason also it is necessary to have the meter revolving (at least approximately) at the same rate. In order to obtain a recording apparatus at a reason- able price, it was necessary to invent a new recording galvanometer which would avoid existing patented articles, and, thanks to the fact that a firm of chemical manufacturers required an automatic recording method for carbon monoxide in hydrogen, the writer was enabled to work out these processes. The recording galvanometer is comparatively simple. A piece of platinum wire bent twice at right angles is fixed with wax to the end of the galvanometer needle. Above the needle a cylinder is placed, and this is revolved by clockwork. The cylinder has on it a projection placed diagonally across it and extending over one-third of the total circumference. This projection is made of metal covered with a piece of velvet. Normally the needle swings free of the cylinder, and it is only when this projection is at the needle that it depresses it (the needle), and the pla’tinum is depressed into two parallel troughs of mercury. This contact of the platinum with the two mercury troughs completes an electric relay circuit, and brings an electro-magnet into action, which pulls a pen on to a disc of paper and marks it. The disc of paper is on the face of a clock which revolves once in 24 hours. The pen is carried on an eccentric attached to the end of the small drum with the projec- tion previously mentioned, and this eccentric is placed so that when the diagonal velvet projection engages and depresses the needle into the mercury at maximum deflection (corresponding to the maximum amount of carbon monoxide to be estimated), the pen is pulled over on to the paper and marks it at the maximum per- centage of the gas; and, vice versa, when the diagonal projection engages' and depresses the needle at no deflection, the pen is pulled over and marks the paper at zero. For intermediate positions of the needle of the galvanometer the pen is pulled over at correspond- ing positions on the paper. In addition to the written record, the apparatus can be arranged (and is preferably so) to bring in an electric bell for a certain range, a second bell of different tone for a higher range, and a third bell for a still higher range. The number need not, of course, be limited to three bells, as there is no difficulty in bringing in any reasonable number. Part II. A mixture of hydrogen with oxygen or air has an ignition point of about 1200 degs. Fahr. (650 degs. Cent.), but if passed over palladium warmed to 140 degs. Fahr. (60 degs. Cent.) combination takes place rapidly. Such a material is called a catalyst. It has up to now been generally supposed that carbon monoxide is more or less a poison to most catalysts without, however, any experimental evidence that the writer has been able to find. During some experiments the action of carbon monoxide on the palladium black was tried, and, much to the writer’s surprise, the sup- posed poison for catalysts was so active in presence of air that it heated up the palladium black from the ordi- nary temperature to redness and ignited. It was found on further investigation that platinum black would act on carbon monoxide in air at ordinary temperatures, and convert the poison into carbon dioxide. There was no doubt about the method of estimating CO in future in mining practice, but could the new discovery be made of real importance for the basis of a respirator, and perhaps be the means of saving many lives which are lost every year by slow poisoning by carbon monoxide? There is no doubt that a respirator can be made which will convert the poisonous carbon monoxide into C02 at the ordinary temperature, the only doubt is—can it be done at a cost which is not prohibitive? The author has not made very much headway with the use of these active metals for the removal of carbon monoxide or for the estimation of it. The preliminary experiments have taken much time, and the results are not ready for publication. Moreover, on account of the high price a very extensive search was made to see if some cheaper material could not be found. It was rather a hopeless task, and up to now nothing has been found that can take the place of these rare metals. After an explosion of firedamp and coal dust with insufficient air carbon monoxide is always present, and is accompanied by hydrogen 2CH4 + 02 = 4H2 + 2C0, for the oxygen always attaches itself to the carbon part of a hydrocarbon first during combustion. The hydrogen is also burnt by these rare metals at ordinary temperature which would help in removing the CO by giving heat. It has been found that the rate at which they act on carbon monoxide in air rapidly increases with only a moderate rise in temperature. It is pos- sible that other gases which may be present after an explosion may poison the catalysts. All these points have to be carefully investigated before it can be decided that it is feasible to make a respirator weigh- ing a pound or so which will be quite certain in its action, and not be too costly. During the experiments we have on several occasions made the platinum and palladium blacks and found them to be inactive at the ordinary temperature, and the cause had to be traced so that it could be avoided in future. PERMITTED EXPLOSIVES. The following is a complete list of the explosives now contained in the Permitted List :— Part 1. (Explosives which have passed the Rotherham Test.} Permissible Pendulum Explosive. maximum charge swing in in ozs. inches.* Abelite No. 1 14 2’85 Ajax Powder 12 2’69 Ammonite No. 2... 10 1’99 Ammonite No. 3 ... 12 2’12 Ammonite No. 4 30 1’76 Anchorite 14 2’73 A 1 Monobel 28 2’78 Arkite No. 2 40 2’41 Bellite No. 2 32 2’42 Bellite No. 4 18 2’92 Britonite No. 2 24 2’26 C ambrite 30 1’98 Denaby Powder ... ... 18 2’74 Dreadnought Powder ... 32 2’05 Duxite 12 2’45 Dynobel ... 22 2’61 Essex Powder 38 2’17 Expedite 32 2’62 Faversham Powder No. 2 24 2’61 ' Haylite No. 1 10 2’18 Kentite 18 2’64 Kent Powder 32 2’01 Kynarkite 20 2’21 Melling Powder 12 2’62 Monarkite 26 2’67 Monobel No. 1 10 2’81 Negro Powder No. 2 ... 20 2’21 Neonal 16 2’56 New Fortex 10 2’61 Nitro-Densite 28 1’47 Permon Powder 18 2’57 Pit-ite No. 2 32 2’15 Pitsea Powder No. 2 ... 8 2’64 Roburite No. 4 18 2’86 Sunderite 16 2’66 Super-Curtisite 16 2’71 Super-Excelite 10 2’74 Super-Excelite No. 2 ... 14 2’72 Superite ’. 10 2’53 Super-Kolax 30 2’10 Super-Kolax No. 2 32 2’21 Swale Powder 20 2’50 Syndite 40 2’22 Tutol No. 2 22 2’11 Victor Powder 18 2’96 .Westfalite No. 3 12 2’55 * This is the swing given to the s ballistic pendulum at the Home Office Testing Station by firing at it a shot of 4 ozs. of the explosive. It may be compared with the swing of 3’27 in. given by a shot of 4 ozs. of gelignite containing 60 per cent, of mtro-glycerme. Part 2. Bobbinite. (Permitted only for the purpose of bringing down coal in certain mines for a period of 5 years from January 1, 1914.) Immingham Coal Exports. — The coal exports from Immingham during the week ended the 11th inst. were comparatively slack, the official returns of the shipments being :—Foreign : To Cronstadt, 2,561 tons; Fernando Po, 3,899; Flensborg, 1,757; Hamburg, 3,951; Oxelosund, 1,710; and Reval, 2,123—total, 16,011 tons. For the Lerwick, 1,300; London, 2,450—total, 3,750. For the corresponding week last year the figures were 23,627 tons foreign and 1,039 tons coastal. Grimsby Coal Exports. — The exports of coal from Grimsby during the week end§d Thursday, June 11, were shown by the official returns to be as follow :—Foreign : To Aarhus, 1,545 tons; Dieppe, 756; Gothenburg, 228; Ham- burg, 1,015; Landscrona, 1,558\;Malmo, 1,005; Skien, 103; Trelleborg, 1,791; Ystrad, 2,?271—total, 10,272. Coastal: To London, 96; Gravesend, 370; Lerwick, 351; and Rye, 180—total, 997. For the corresponding week last year the totals were 14,763 and 1,661 respectively.