Investigations of the Removal of the Toxic Gases, Arsine and Phosgene, from Gas-Air Mixtures by Impregnated Charcoals, Using the Radioactive Indicator Method Citation Martin, Don Stanley (1944) Investigations of the Removal of the Toxic Gases, Arsine and Phosgene, from Gas-Air Mixtures by Impregnated Charcoals, Using the Radioactive Indicator Method. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/J4FA-ZA04. https://resolver.caltech.edu/CaltechETD:etd-12192007-093234 Abstract No abstract submitted. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Yost, Don M. Thesis Committee: Unknown, Unknown Defense Date: 1 January 1944 Record Number: CaltechETD:etd-12192007-093234 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-12192007-093234 DOI: 10.7907/J4FA-ZA04 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 5068 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 19 Dec 2007 Last Modified: 10 Nov 2023 19:32 Thesis Files Preview PDF (Martin_ds_1944.pdf) - Final Version See Usage Policy. 71MB Repository Staff Only: item control page

Investigations of the Removal of the Toxic Gases, Arsine and Phosgene, from Gas—-Air Mixtures by Impregnated Charcoals, using the Radioactive Indicator Method Thesis by Den 5. Martin, Jr. In Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 19d TABLE OF CONTENTS Acknowledgment 2... ccc cc ccc cccnccercsevecesvesverecessses dL Introduction ...ccccscccccsccccncccrenccasccccssesecesess 3 Mathematical Introduction ....c.scsccercccscccscccccenene GF Definition of Symbols wcccccsscccrecsevesecseces General Considerations for Tube Tests ....eese++ 10 Case 1 scccccccccesccesesscccsccccccencesessceas Le CASE 2 ceccserncnccccccccccccnceccceesscseceeces LG Other Cases .cccccccncnccsscctccecccscccsesesees Ob General Kxperimental Procedures and Materials .....eese06 23 Charcoal .cccccccccccvcccscovccsccccccsecccesees OI Procedures for Tube Tests ...ccccececccccececces 2h Preparation of AimGas Mixtures ..cccccccscesees 2h Temperature Measurements .....cccccccsecsccscese 29 Radioactivity Measurements .....cecsccccccccsces 20 Corrections and Errors ..ssccesccccrescccsecsees 29 Preparation of Graphs ..ccccecsesccccccecccssces Sh Experiments with Hadioarsine ...sessscccccesccscccscssces SF Preparation and Separation of hadioarsenic and Synthesis of Arsine ......sscesceeee 33 ANALYSES cecccccsccccenccensenscessascseccsensecs 35 FPube Tests ..ccecccccccccccasescccccacccoseseces 30 Accumulation and Distribution Experiments ...... 39 Experiments with Hadiophoscene .cccsccccsececcncececsces Preparation of Inactive Phosgene cesesccccccves Preparation of Phosgene Containing Active Carbon Sevccecvccecesscearscesesees Preperation of Phosgene frou Active Chlorine .. Chemical Analyses and Tests for Phosgene ..ee.. Tegtes with Radiocarbon cvcoccecesvesuevenssveee experiments with Radiochloringe cecscsccsscceves Thermocouple Measurements .cccsscsececssceeeses Discussion of Results .ecsccnescrccccasvcsonces Summary SHEESH SFA HHHPHEESH HEB O oT HA SEHF GOH HRHEHKEC SHEATH HERE Bibliography PASFG HFFA HSHKHSSESEFSHEEFESEHEHEHESERERHEHRESTEHE EHO R ES Appendix 1 Graphs for Arsine cxperisents Appendix IT “Graphs for Phosgene sxperiments with Radiocarbon Appendix III Graphs for Phosgene Sxperiments with Radiochlorine Aopendix IV Graphs for Temperature Measurements in Phosgene Tests 45 45 oe) 46 47 The author desires to express his gratification for the stimulating assistance and guidance of Dr. Don M. Yost, under whose immediate direction these studies were performed. The association with Dr. R. We Dodson in the experimental work is considered to have been invaluable and was, indeed, very pleasant. The assistance of Dr. E. 0. Lawrence and the staff of the Radia- tion Laboratory together with the hospitality of the Departments of Chemistry and Physics at the University of California at Berkeley was greatly appreciated. The aid of Dr. Martin Kamen in preparing radio- active samples was very helpful. The work described in this thesis was performed during parts of 1941 and 1942 in collaboration with Dr. R. W. Dedson and directed by Professor Don M. Yost under the National Defense Research Committee Contract B 102, Section B6, OEMsr-28, with the California Institute of Technology. The results have been presented previously in reports to the N.D.R.C. (1, 2, 3). This research was undertaken to study urgent chemical warfare problems which had arisen and actually represents only a emall fraction of the intensive work in this country that has been devoted to furnishing satisfactory protection against the toxic chemical agents of battle. The experiments, although fundamental in nature, were planned with the specific practical objective of affording an understanding and a correlation of the processes actually occurring in gas mask ab- sorbents in order that it might be possible to design better canisters. Possibilities for further interesting research in this field were indicated by the results obtained, but the urgency of other war problems prevented the logical extension of this program. Since this thesis deals with research and development done in cooperation with the armed services, it must, of course, be classified as "confidential." Because a large portion of the scientific litera- ture dealing with the subjects of interest exists as classified reports which were not available to the author, the bibliography is un- fortunately severely Limited. It is earnestly hoped that the work reported has contributed significantly te a satisfactory chemical warfare development program. The use of chemical agents on a large scale in battle and the spectacular results of the "gas" attacks that occurred in the First World War are well known. Highly effective chemical warfare tactics were devised whereby substances extremely toxic to humans were dis~ persed over the battle fields to harass, injure, and kill the enemy. According to Prentiss (4) the most extensively used class of agents and the class accounting for the greatest number of fatalities con- sisted of the nonpersistent "gases." These substances had relatively smal] molecular weights and low boiling points; so when released, they quickly vaporized and mixed with the air. Even wit considerable dilution the quantities that were drawn into the lungs of the person- nel with the inspired air and absorbed were capable of producing serious casualties. That these newly invented chemical tactics did not determine the course of the war was due partially, at least, to the development of adequate protective measures. Protection against the nonpersistent gases was afforded by gas masks in which the inspired gases passed through a canister of suit~ able absorbents that removed the toxic constituents. Charcoal has been found to be the best practical absorbent for these purposes; all the present day combat gas masks employ a bed of granules of some charcoal material. The original char, which may come from various sources, is usually activated by suitable oxidation processes and may be impregnated with substances which greatly enhance its absorbing powers. Recipes for producing the adsorbents are different in every country and are being constantly modified to yield improved products. The weight and size of a canister which the troops carry on their person must be kept to an absolute minimum consistent with "adequate protection", and a very thorough and careful design of the gae mask is essential. For proper design the most nearly cosplete information possible concerning the kinetics of absorption and the processes occurring have proved imperative. Final testing of canisters mist, of course, be performed with the war gases at concentrations that can be encountered in the field in actual or simulated breathing tests. For purposes of economy, however, a preliminary development program must be based on a large number of tube tests. In such tube testa the air-gaa mixtures may be passed at a constant rate into a bed of the charcoal absorbent supported in a small tube. Either physical or chemical means may be utilized to detect or measure quantitatively the amount of toxic material remaining in the bed or leaving with the effluent gases. The data which can be gained from the tube experiments may then be analyzed to yield information con- cerning the kinetics of the absorption. The correlation of these kinetics studies to actual canister behavior may follow; although it has been found that considerable care must be exercised in extending the test results to properly account for maximum flow rates and the usual shallow bed thicknesa in canisters. If tube tests are analyzed extensively by chemical methods to determine charcoal performance, a large number of quantitative analyses may be involved. The value of physical methods by means of which the concentrations can be measured at frequent intervals,or perhaps be automatically recorded, can be well appreciated. Yost, Dodson, and Cooley under a previous N.D.R.C. contract (5) developed techniques and constructed instruments for determining the relative intensity of radiation emitted per wnit volume by a radio- active substance contained in a flowing gas stream. Counters of the Geiger-Mueller type for measuring essentially the beta radiation were employed in conjunction with ratemeter type electric circuits. The ratemeter produced a current through a milliammeter which was directly proportional to the counting rate of the counter tube. By the use of recording meters a continuous record of the radioactivity could be obtained. After the proper correction for radioactive decay, the activity per unit volume of a gas, in which a single radioelement was . present, would be essentially proportional the concentration of that particular element in the gas with an accuracy determined by the in- herent statistical fluctuations. It is to be especially noted that the concentration of the tracer element would be obtained, regardless of the various molecular states in which it might exist. The chemical properties and behavior of the compounds containing radioactive iso- topes are not different in any important respect from those of the same compounds containing only non-radioactive isotopes. The use of the above method for determining concentrations was applied in studies of the absorption in tube tests of the two gases, arsine and phosgene, as described in this thesis. The radioactive indicator technique permitted a number of interesting experiments. It was possible to obtain continuous records of concentrations in the effluent gases from the charcoal tubes. The identification of chemical compounds could be accomplished by selective removal of the components from the gas stream and the subsequent measure of the residual activity. Also, it was found feasible to measure the quantity of radioactivity deposited in the charcoal as a function of either time or the position in the bed. In addition to the naturally occurring radioactivity there have been prepared and identified in recent years a large number of arti- ficial radioactive isotopes, including those of nearly every element. A table of the known radioisotopes has been prepared by Seaborg (6). For charcoal tube tests it was necessary to have available an un- stable isotope of st least one of the elements of the toxic compound with a half life sufficiently long so the activity could still be accurately measured at the end of a two or three hour experiment. Rather high activities of these isotopes were required. As small a fraction as 107* of the original radioactive sample was needed for measurement. If a concentration of 0.1% of the initial concentration was to be detected as 10 counts/min., about one half of a low back- ground, and the efficiency of counting was 5%, which was essentially a geometry factor for energetic beta radiation, then an original activity of at least the order of magnitude of 2 « 107 counts/min. or about a millicurie was required. (One millicurie may be defined as 2.22 x 10? disintegrations per minute.) The larger cyclotrons constitute the only present day source of activities of this magnitude. Beta radiation with energies of at least 1 MEV. was the most convenient for measuring; although activities for © particles with energies as low as 0.1 MEV. could be obtained. An efficient and rapid synthesis, requiring not more than one or two half lives, of the compound to be ao used from the target materials of the bombardment was essential. It is not possible to prepare an isotope of some elements with a suitable decay constant and in quantities sufficient for tube test experiments; the method is somewhat limited for this reason. In case satisfactory isotopes could be obtained for more than one of the elements of a compound of interest, the behavior of each of these elements could be observed. Arsine, AsHs, is a colorless gas with an unpleasant odor and condensing at -55°C. It is a systemic poison, extremely toxic; and the inhalation of small amounts can result in death several days after exposure, according to Henderson and Haggard (7). Its value as a war gas is doubtful in view of the ease of its oxidation and decomposition and its low boiling point. It may, however, be considered as an example of a gas that can be easily oxidized on the catalytic charcoal surface. Arsine may be prepared from radioarsenic; the isotope that can be con- veniently utilized is as’*, the radiation is @, @, and ¥. The beta radiation is quite energetic, 0.9 MEV. for Cand 1.3 MEV. for @ (6). Its half life, 17 days, is moderately long, so the tube tests may be conveniently carried out. The radiation from radiohydrogen is so feeble, 0.02 MEV. (6), that it could not be used. Phosgene, COCl2, is a colorless gas with the odor of mouldy hay; it boils at @°C. it is primarily a lung irritant of high toxi- city, and small dosages may result in fatalities as long as 24 hours after exposure (4). In the presence of water it may hydrolyze on the charcoal surface to give the acid products HCl and COz.. Since it can be cheaply prepared, it has been one of the most widely used and effective of the chemical warfare agents. The carbon isotope cit is available. It emits @" radiation, 0.95 MEV. (6), half life 20.5 minutes (8) which is satisfactory for use. Also, radiochlorine, o1 78, @” radiation of 1.1 and 5.0 MEV. and Y¥ of 1.65 and 2.15 MEV., half life 37 min. (6), can be used. Both its beta and gamma radiation are extremely penetrating. The known unstable isotopes of oxygen, ots and 019, have half lives of 126 sec. and 31 sec. respectively; these are too short to be useful. Indsed, the half life of 20.5 min. for radiocarbon represents very nearly the lower limit that could be utilized for experiments of this sort. MATHEMATICAL INTRODUCTION Theoretical considerations have yielded several mathematical equations for describing the results of tube tests. The aim of a theoretical treatment is the definition of a relatively few para- meters, such as absorbtion capacity and absorption rate constants , which will characterize the worth of the individual charcoals. The concentration of material in the gas stream or retained on the sur face of the charcoal may possibly then be calculated as functions of the time and the position in the bed, these functions containing the assumed parameters. The actual concentrations can then be compared with the calculated values. If the calculations are satisfactory, values of the parameters can be obtained from s small. number of ex- periments. The performance of a charcoal under different conditions or in actual gas mask use may then be calculated. Some mathematical treatment of the problem is essential if any correlation between ex- periments is to be obtained because of the large number of variables and the complexities involved. When the absorption studies were started, there were available in N.D.R.C. reports only a few very fragmentary discussions (7,8,9) of the mathematical expressions that pertained. The lack of suit- able system for designating reports by the N.D.R.C. at that time prevents satisfactory references to these from being given. The following mathematical treatment, guided by the available information in the reports, was accordingly worked out for the purpose of inter preting the tube tests with radioactive indicators. ~10~ Definition of Symbols. The following symbols may be defined. A = cross-section area of the tube (cm.“) ¢ = concentration of gas in air stream at x and t (mg./1.) Cp* influent concentration of gae (mg./1.) L = length of absorbent bed (cm. ) N = concentration of "gas" absorbed in the bed, mg "gas"/cc, absorbent t = time (min.) t= inlet time for a particular increment of the air stream (min.) . x = distance of a point in the bed from the inlet end (cm.) v = linear velocity of flow within the bed (cm./min.) V¥ = volumetric rate of flow (1./min.) General Considerations for Tube Tests. A cylindrical tube, packed with homogeneous absorbent, through which flows an air stream that carries an absorbable gas was considered. The concentration of gas was assumed to be so small that no appreciable change in the linear velocity occurs because of its removal. It is possible to set up a general differential equation from a mass balance over an infinitesimal section of the absorbent bed. \v =Co WoNion | 4 c2C(Xt), NEN 4 ax L i Fig. A i ACO C=C(L, 1) \v In the portion of the bed between x and x + dx, indicated in Fig. A, the rate of removal of material from the air stream with respect to time would be -v [0 (x + dx, t+ dx) - ec, tI. In the expanded form, neglecting differentials of order higher than one, this becomes . . -¥ | etx, t) G2), + (52) a - e(x, o| The rate at which material appears on the absorbent is Adx (28 dt’, Since the rate of removal of gas from the air stream must equal its appearance in the bed, the above expressions are equal; thus x A considerable simplification may be achieved if the independent variables, x and t, are transformed in the following way. xX*®X 35 t. =t- ; By means of this transformation equation 1 is changed to Ge) = 4 G8) at x A \gx/,' . Equation 2 contains no mechanism considerations but is only a state- ment of the Law of Conservation of Matter. Hence it is general and applies to any tube test experiment. The quantity, %, is ordinarily very small, of the order of Q.1 sec. Therefore t’ can be treated essentially the same as the variable, t. The statement of the kinetics or mechaniam may then -]2- ON comprise a differential equation in which (3 t) is expressed as a function of c and N. Upon solving this equation and equation 2 simultaneously in accord with the boundary conditions, N and ¢ will be obtained as functions of t and x. Case 1. A useful and convenient solution is obtained in case the charcoal is assumed to have a saturation capacity, N,; and if the rate of pick-up by the charcoal is assumed to be proportional to the concentration of gas in the air stream and to the unfilled capacity of the charcoal at any point and instant in question. The differential equation applies (5) ~ Me Gos) L ae (3) (The usual convention has been to call the unfilled capacity of the charcoal Ne Equation 3 would then be é k, oN. (S$) + * (4) x However the concept of saturation capacity is definitely a mechaniam assumption. By choosing to designate by the symbol, N, the capacity filled with toxic, the generality of equation 2 is enhanced.) Equations 2 and 3 may be solved for the boundary conditions which corrsspond to normal tube testing. at t' = 0,N=0 forO<x <b Ab x =0,¢=¢9 fort >O From equations 2 and 3 there obtains -v [a - | t (3g), = GP 6 -13~ Differentiating equation 4 with respect to t’ and eliminating (NN) ana (SE ) x >Y means of equations 2 and 5, there results upon rearrangement the second order partial differential equation which does not include N. | The expression on the left hand side is a perfect differential with respect to x and consequently can be integrated immediately to piesld 3 (58) racer (7) where f (t’) 4s a function of t’ only. From the boundary condition, ¢ = 0, for t’> 6 at x = 0, it appears thet f (t/) = key, Using this value for f (t') and integrating for constant x gives In (S972) . Kyeot! + a(x) (8) c where g(x) is a function of only x. learranging terms to obtain an explicit expression for ¢ gives ket c= _soe 7° (9) k ' [ 1Sot ob *) | But if this value for c is substituted in equation 3 and integrated for constant x from t’ = 0 to t! = t, the solution involving N is obtained in & ~N } an + oe) | (10) No e kicot’ . g(x) Thus, solving for ¥,-N and W (x) ' e1°o", 98 (*))’ 3° ; Kiet”, ° E(x) (11) There remains now only to identify g(x). A differential equation for g(x) is obtained by writing equation 5 in terms of t and g(x) $2) - wee tot SB g B(x) dx [east yee) J k,¢,.% ~ hytelNo “NL A Sollee _ la+e e@) | - (12) v Vv. f kyegt’ g(x) | 2 @ +e Solving for g(x) gives Higley Ax Vv g(x) = In | ne -2 | (13) where I, is a constant of the integration. I,, evaluated from the boundary condition, is 1. The complete solution for Case ] may now be written Noky Ax e -1 e Vv an (2-1) “-kKet + 1n (1ha) or alternatively , c= 2 wo Roki (4b) [ o l+e V | and é k et N=N 8 es ; (15) ° [ ket" , Nok {Ax ® e v ~ 1 ~15~ The equation lia has been especially useful. As far as is known, it was apparently first applied in this war by the British; and it has been given the name, Hinshelwood's equation. The details of Hinshelwood's treatment have not been available; however the | equation in the form of lib was used in the last war. A derivation by Bohart and Adams (9), who assumed equations of the form 4 and 5 in the variables x and t, was published in 1920. If for a single tube test the toxic concentration in the ef- fluent is measured as a function of time, ecuation lha predicts that the function, in( <2 - i) , plotted against the time will be a straight line with the slope, wk, es Such a procedure may be used to test ex~ perimentally the walidity of the original kinetics assumptions and the resulting equations. If approximately a straight line is obtained, then numerical values for k, and uy may be calculated for the parti- cular gas and charcoal from a single experiment. It also appears from Hinshelwood's equation that toxic material penetrates the bed in increasing amounts after the start of the test at t «0. In practice a break is considered to have occurred when some detectable concen- tration of gas, ¢,, penetrates the bed. If the time, at which this concentration is cbserved in the effluent, is designated as the life, , of the bed, then for a bed of length, L, the equation may be written , s ¢ oe Oe ee _). (16) e ~~ If =r" is sufficiently large, then c ‘ , Thus, as L increases, the life-thickness, ty ~ L, curve should approach a straight line, From the slope of this line and from its intercept both ky and N, may be calculated provided the value of o, is known. Case 2. In the derivation of Case 1 it was found that N could be eliminated between equations 2 and 3 and the resulting differential - equation of second order for ¢ could be integrated immediately. In general, the integration can not be performed so easily. It was found however that a satisfactory solution could be obtained for the case in which the rate of absorption is proportional to only the gas con- centration in the air stream for a region where the absorbent is not saturated. Such a condition might be approximated if the rate de- termining step was the transport of the gas to the charcoal surface. In place of equation 3 in Case 1, the two kinetics equations may be written (23,) = kee for NSN, , (18a) x oN It". = OforN=N, . (18b) x The same boundary conditions as for Case 1 apply, viz. Att =0,N=0 for 0 <x <L, At x = 0,c¢ 6, for t’>0 By combining equations 2 and 18a there is obtained ac kgAc <2) ee ee ee (19) t This may be integrated ~ kA “i x (20) c= fy(t') e From the boundary condition, ¢ = ¢, for t'> Oatx= O, it appears that f(t’) = 5; 80 keA Tx c= ce . for N <N.. - (20a) Substituting this value of ¢ in equation 18a and integrating for constant x front = 0 tot’ = t’, the expression for N results keA 7 nnnann v N= ketot @ for N <4. (22) Equations 20 and 21 apply during the early portions of the test be- fore any charcoal has been saturated. During this period the gas concentration in the air stream is independent of time; and, con- sequently, a constant small concentration appears in the effluent. The amount of gas absorbed at any point of the bed increases linearly with time. At x = 0, N reaches the saturation value, No», in the time tos which ie equal to N,/k2e,. This ty may be considered as a "theo- retical" break time. For at t, the top layer of the chercoel has be~ come saturated. A zone of saturation then begins to expand from x = 0 through the bed; simultaneously, the effluent concentration begins to increase. Then ¢ is no longer independent of time. If x, is the thickness of the saturated zone at time t’, x, will be a function of t'. Equation 20a will still be valid if x-x,(t’) is substituted in place of x, since the equation is valid for any region in which N <¢N,. Thus, keA i" ry] --" a- x,(t ) , ' N c= oe ° for t >toree , (22) ~13— Patie kad N= Noe v* . (23) The time at which the edge of the saturation zone reaches the point, x, in the bed may be called (x). The rate at which the boundary of the saturated zone moves through the bed is then / aT’. But x dx this is also equal to “"o . Thus, dt! , yit. %, (24) (—e " OR If the quantity, ( a) » is integrated for constant x fromt’ = ty dt’ x tot’ = T', the following results 1 " keh (22)) dt = Ny (... v . (25) +! This equation may be differentiated with respect to x to give vr! ad a ON ' o*N ' aN ' at’ ax Gt) at jxpu + [ser x) ax t t! x _ KoA = “oe e V (26) at at) x equation 2 and 18a respectively and the velue of the reaulting de- ; z But if the expression for /2.4 ) and .9.N_ be substituted from x finite integral taken from equation 25, the above equation becomes -18- —* (23) The time at which the edge of the saturation zone reaches the point, / x, in the bed may be called (x). The rate at which the boundary 4 of the saturated zone moves through the bed is then 1/ aT . But dx this is also equal to %o . thus, at! 1/ aT. (24) dx qb pea If the quantity, (2: ) » is integrated for constant x from to= ty ’ jo ct x tot = 7 , the following results "Vy koA on toy -_e Vv (25) (2 ) dt No ( l-e . x by This equation may be differentiated with respect to x to give t T T' d AN) a! d7n 1 [on yt at’ 1 v bo vo ° _ keA x = Nok. A e v (26) Vv e - 2 But if the expression for aN and dk be substituted from Qt’ dtdx equation 2 and ]8a respectively and the value of the resulting de- finite integral taken from equation 25, the above eguation becomes when combined with equation 24 dt’ , Veo _ aX 1/ o— ine wT: (27) The saturation zone would appear to move through the charcoal bed, its length increasing linearly with time, Equation 27 may therefore be integrated at once to give Ve = t' =! (28) Placing this value of x, (t) in equation 22 yields the explicit expression for ¢ Tapo tape (tt c= Ge ° 0) for ty < t'<tg + (29) At The similar expression for N may be obtained by integrating si, ’ x which 4s equal to k.c, fromt'=t' tot’ = T', using the value of ¢ given in equation (29). Thus, v! tr" KpA ok ‘Qu ' , - ==; o (t) - to) ; ati) dt =Ng-N- Katy e@ “o dt t’ x t' ' _ kak coke | T v * NS ft Or to) ay N,e@ Qo (30) +! From this it follows _ + > (te - t.) tor t , é ! é t! ’ AN5x (31) = @ : ° OF Lo o* “Veo The results may now be summarized by the get of six equations hey! a le ~ ce For 0<t St, Kaeo 1 @ = Ge (22) ~20~ ' AN V N ™ %o For t, Ct A ty + Fm ro = age ° ; (29) ’ ’ AN Fort = tot ee eno, (32) __ kgAx Le 2 ' No ’ v For QO <¢t = to “Emon t = keegt @ . (21) gba, . ' 1 or AN V N = %o For t,<t<t, + ~ :N = Noe (32) fy ot AH For t = ¢ tN aN 33 © Cy fs) (33) Although some toxic material penetrates the bed immediately, the effluent concentration would remain constant until the "theoretical" break point, t.. This behavior is in contrast to Case 1 where the effluent concentration increased continuously. The "theoretical" break point for any one charcoal against any particular gas would be inversely proportional to the influent concentration, ¢,, and would be independent of the bed length. It might be, however, that for the early part of the test the effluent concentration in the case of a sufficiently deep bed would escape detection. In which case equation 29 would apply for the practical break time. Equation 29 may be written in the logarithmic form for a bed length of L i] § in se _. Soks(t oo, Kohl. tor t') t) (3a) Q in this case the function ln =) would be a straight Line function of t > and the validity of the mechanism could be tested by plotting the experimentally determined 1n <2 against time. As with Hinshel- wood's equation the parameters may Le determined from such a graph if the results are suitable. It appears also that the life thickness curve would be a straight line for sufficiently deep beds. The parameters, kg and N,, could be evaluated from a satisfactory life-thickness curve. As the bed length was decreased, the break time, th» for some detectable concentration would appreach t,. When by the above equation L was so small that t, < t,, the bresk time would be sero. The two theories that have been presented may aid in evaluating charcoals. If the equations are approximately valid, the parameters, k and N,, offer a convenient means for comparing different charcoals, The best absorbents will have high values for both Ny and k. Ifa charcoal possesses a high value of No and a low value of k, it may provide satisfactory protection in deep beds and fail when used in thin layers. Of the two cases discussed, Case 1, yielding Hinshel- wood's equation, is probably the most valid for the present gas mask absorbents. Qther Cases. The case of desorption has not been discussed, although many of the absorption processes are reversible. Kassel (10) considered the rate equation (34) = Hold + Keto = (35) in which kg would be a desorption rate constant. A solution in the form of a complicated series was given for this case. Only incomplete numerical solutions were calculated, and because of its complexity these solutions have not proved useful. it is apparent that stil] other mechanism differential equations could be set up. In some instances, at least, the integration for Dem such other cases will be difficult; and simple expressions for c and N as functions of x and t cannot be obtained. “23+ GENERAL EXPERIMENTAL PROCEDURES AND MATERIALS Charcoal. The absorption studies were made with the gas mask charcoal designated by the Chemical Warfare Service as CWSN19TUS. The unimpregnated or base char for this whetlerite, CWSN19, was also used in a few experiments. The base char, an extruded material, had been made from wood flour by a zinc chloride method, performed by the National Carbon Co. The National Carbon process formed a sizable fraction of the production capacity for charcoal. in this country, and subsequent testing has proved it to be undersirable for gas mask use. The base char had been given a cupric ammonia carbonate impregnation followed by a high temperature air drying so that CGWSN1OTUS was a whetlerite that would now be classed as type A. It therefore contained copper, about 5 to Th by weight, in the form of Cu0. Its surface would be an active cata- lyst for many reactions, and its behavior might be axpected to be con- siderably different from an unimpregnated activated charcoal. All the charcoal used was screened with standard sieves, the fraction passing through 10 mesh (per inch) and over 1/, mesh being selected for tube tests. This fraction was dried at least four hours at 150°C. It will be designated as CWSNLOTUS (Dried). ‘Two batches of the dried whet- lerite were equilibrated at room temperature with air at 50% and 70% relative humidity (RH) respectively. These portions will be designated CWSNL9TUS (50 RH) and CWSNI9TUS (70 RH). The equilibration was con- tinued until the water pick-up over a period of several hours was negli- gible as indicated by repeated weighings. The 50 RH material was found to contain 7.0% Ha0, the 70 RH charcoal 27% H2Q, on a wet basis. Zhan Procedures for Tube Tests. For tube tests the charcoal was con- tained in either aluminum or glass tubes (ca.2cm.1D) and was supported by Witt plates. The aluminum tubes had thin walls (0.04 cm.) and were a@specially designed for the purpose of permitting the accumulation of activity in the bed to be measured by Gelger—Mueller counter tubes. It was found necessary to coat the inside aluminum walls with shellac to prevent corrosion. Experiments at room temperature were performed with the tubes exposed to the laboratory atmosphere. In other cases, however, the tubes were immersed in liquid baths. Above room temperature a water bath was heated with a hot plate; below room temperature an ace- tone bath was cooled with dry ice. In each experiment a weighed amount of charcoal was introduced in- to a tube; and the bed depth was adjusted to a predetermined value s0 the apparent density of dried whetlerite was a constant, and in the ex= periments equal to 0.50 gm./ec. In the case of an equilibrated char- coal the weight of absorbent used was made equal to the weight of dried charcoal in the adjusted bed depth plus the weight of water absorbed by this amount of dried material in the equilibration process. The tubes were always mounted with their axes in a vertical position, and the air~gas stream was passed down through the bed. The effluent stream was frequently passed through chemical indicator solutions in test tubes for the detection of the toxic constituent and to determine standard chemical break points. Preparation of Air-Gas Mixtures. Ordinary steel oxygen cylinders (Volume ~ 23L) were prepared. Their interiors were carefully cleaned, dried, given a coat of shellac, and finally dried again with a current ~25= of hot air for several hours. To fill a tank, it was first evacuated by a mechanical pump. A glass bulb (about 1 to 2L capacity) filled with @ sample of the radioactive gas was attached to the evacuated cylinder. The sample of gas was drawn into the cylinder; then dried air was flushed through the bulb into the still partially evacuated tank. If necessary, a sample of inactive gas from another bulb could also be introduced in the same manner. Finally, compressed air from another steel cylinder was adwitted until the pressure of the air-gas sample was 500-600lbs/in, gage. The cylinder from which compressed air was drawn had not been initially dried. Subsequent examination revealed that it contained a gmall quantity of liquid water. In use, the pressure of these tanks was always greater than 1000]bs/in*, Air, saturated with water at these pressures, would have a relative humidity of less than 2% when expanded to one atmosphere. ‘This figure may be considered an upper limit to the relative humidity of the gas-air mixtures. The cylinder containing the sample was provided with a regulator and a needle valve, these valves being used to control the flow from the tanks. Conventional flow meters of various ranges were employed to measure the rates of flow, and barostats were used to secure uniform air velocities. If needed, air from a compressed air line could be mixed at a controlled rate with the gas from the cylinder to provide further dilution. Desorption was studied by passing air through a char- coal after the air-gas mixture. A portion of every sample in the cylinder was withdrawn for quantitative chemicsl analyses. Temperature Measurements. In early tests it was noted that the re- moval of the gas was accompanied by thermal effects. Therefore in later studies thermocouple junctions, made from silk covered fine copper and eonstantan wire, were inserted in the charcoal at various distances along the bed so temperatures might be measured. The potentials were measured with a Leeds and Northrup portable precision potentiometer. Radioactivity Measurements. The counter tubes and ratemeter circuits, that were used for measuring radioactivities, were constructed under a previous N.D.R.C. contract and have been described in a formal report by Yost, Dodson and Cooley (5). The counters were bell-shaped and of the Geiger—Mueller type. The flat end of each tube was covered with a window of thin mica, the area density of which was 2 to 5 mg/cm*. The counters had plateaus of 200 to 300v. in the neighborhood of 1200v. and backgrounds of 20 cts./min. when shielded with 2 in. of lead. Their excellent caunt- ing characteristics appeared to be stable for long periods of time. As stated previously, the counter tubes were operated in conjunction with ratemeter circuits and milliammeters which automatically recorded a current proportional to the counting rate of the tubes. The ratemeters had seven ranges, the full seale counting rates extending from 200 cts./min. for low activities up to 20,000 cts./min., the maximum range of the instruments. Gas samples were contained in cylindrical glass pots for the measure- ment of their activities. These pots, about 5 cm. in diameter had a volume of approximately 50 ml. Over one end of each had been placed either a mica window similar to those for the counter tubes or else an aluminum foil window, one mil thick. With phosgene it was necessary to coat the inside of the aluminum foil with shellac to prevent the corrosion of the metal and the formation of a film containing active chlorine. The pots were mounted directly under the counters and both units enclosed in a shield of cast lead whose walls were 2 in. thick. The windows of counter and pot were placed adjacent and very nearly in contact. The mounting was so arranged that the position of the pot with respect to the counter was quickly and easily reproduced. Energetic beta radiation from the gas could easily penetrate both windows. For such an arrangement it was es- timated that about 5% of the beta particles from the gas contained in the pot would enter the counter and be recorded. The pote were constructed with delivery tubes so that a stream of gas could be swept through them, in practice usually at a rate of 2 L./min. In this way there was measured effectively the activity per umit volume of the flowing gas stream. For each gas-air mixture the activity of a sample taken directly from the steel cylinder was determined. With this value and with the chemical analysis of the tank mixture, the concentration, ¢, of the radioelement in the gas stream could be calculated from its measured activity, I, by the simple relation | e (gm. atom/unit vol.) = = ne, (mols/unit vol) (36) ° where ¢, is the concentration of toxic material in the mixture from the cylinder, I, the activity measured for the cylinder mixture, and n is the number of atoms of the tracer element per molecule of the toxic com- pound. fo measure the relative activity in the charcoal itself, two methods were used. In the first, the bed was made up in 0.5 em. sections separated by fine brass screens. After passage of the gas stream, either short of or beyond the break, the sections were removed and their separate activities recorded with the counters. The measured activity was an average over the length of the section and was taken as representing the activity of the midpoint of its section. This method could not be used with the radio~ carbon and radiochlorine because of their rapid decay. In the second ~28- more convenient method, the thin walled glass or aluminum charcoal tube was mounted immediately adjacent to a horizontal slit. The slit was formed by two lead bricks beth 10 cm. thick separated by 3 mm. Behind the bricks, its window pointed toward the slit, was placed a shielded counter. The counting rate so determined was an indication of the activity in the chareoal section directly before the slit; but a correction for the scattering of the beta particles by the walls of the slit was necessary. In order to investigate the effects of electron seattering on the measurements with the slit, beds of varying thickness of charcoal con~ taining active arsenic were set up and scanned; ie, the activity was measured with various sections of the charcoal bed in front of the slit. For a given epecifie activity it was found that the measured activity at the center of a section increased appreciably as the section width was increased beyond 3 mm., this effect being attributed to scattering of oblique radiation by the lead walls of the slit. To arrive at a repre- sentation of the average activity of a given 0.5 cm. section of charcoal , the activity of such a section was observed with the slit center in line with the seetion center. Then, additional 0.5 cm. sections of charcoal containing the same specific arsenic activity were placed adjacent to the first section. After each addition the activity at the midpoint of the first section was redetermined; also the entire bed was scanned. The effect of sections added after the second 0.5 cm. inerement was slight. The results of these measurements were represented by the following relation: tT . x) oO , 7 y 13-1, [o. (Tyr * Ty.) + 0-04 (Ta, | : (37) =29— where j is a running index along the bed, i is the activity of a 0.5 cm. section alone, and q, is the activity measured in the presence of adjacent active 0.5 om. sections. This relation was used to correct measurements made with the lead slit system. Tg was taken as representing the concentration of active arsenic at the center of the 4th 0.5 cm, section. Because of the short half lives the scattering corrections for carbon and chlorine were not obtained. With the aid of the above scattering correction it was possible to quickly scan a charceal bed and obtain the distribution of activity there- in, It was also possible to determine the accumulation of activity at any one section during a tube test as a function of time. In this ease the correction for seattering was less certain, but estimates of the correction were made where possible. | Corrections and Errors. In addition to the instrumental and ob- servational errors ordinarily encountered in physical measurements, the accuracy of the data was also dependent on the statistical fluctuations inherent in radioactive decay. This statistical error would vary with the activity of the sample and would be magnified at low counting rates. It would amount to perhaps 25% at the lowest rates during the detection of minimum concentrations but would be reduced to less than 3% at the highest counting rates. | Flow rates, chemical analyses, and other measurements were subject to errors of about 3%, an accuracy considered consistent with the needs of the studies. For each experiment the background for the counter was determined, and this value was subtracted from the recorded counting rates. A -30~ complication to the background corrections was introduced in the experi- nents with radiocarbon and radiochlorine. The activities of these elements were initially so high that the radiation, penetrating the steel cylinder and scattered through openings of the lead counter shield, was able to raise the normal background several fold. The magnitude of this contri- bution at times of interest was estimated by interpolation; however when this effect was large, the activity measured was also great so the per~ eentage correction was small. In certain experiments when activities greater then the maximum range of the ratemeters were encountered, metal absorbers were placed between the pot and counter; the absorption factors were later estimated and applied to the observed rates, These factors might have been in error by 5%. Because of the rapid decay of carbon and chlorine activities it was necessary to calculate all their activities to some reference instant of time; the start of gas flow through the charcoal was selected for this point. Since in many cases the correction to time zero was made over several (as many as ten) half lives, a slight error in the half life would be magnified in the activity calculated for time zero. In correcting the carbon activities, the half life was taken to be 20.5 = 0.1 min., as determined by Yost, Ridenour, and Shinohara (13) and as confirmed by Solomon (8). The value of 37.0 % 0.4 min., given by Van Voorhis (14), was used for the half life of 0138, Because of the conveniently long half life of as’ , no decay correction was needed for tests of two to three hours duration. The ratemeters were calibrated with a pulse generator and readjusted once a week during the course of the experiments; in general the calibration -31- ef each scale was found to have changed by less than 5% during this time. The small drifts in calibration were erratic and seemed to be temperature dependent. An additional factor contributing to errors in the counting measure- ments was high frequency “hash" oecasionally picked up from the power line. This caused apparent counting rates to increase markedly during short time . intervals, The disturbance was minimized by placing a filter across the line; in the few cases when otherwise unaccountable humps appeared in the milliammeter trace, they were disregarded. The records of the milliammeters were corrected for the time lag inherent in the ratemeter circuit. A tank circuit in the ratemeter con- tained a large condenser,’ C, and a parallel resistor, R. A current, Ip , in the form of small pulses for each discharge of the counter tube, was fed into the tank condenser. Since the milliammeter recorded effectively the current, Ip, flowing through the resistor, the current, Ip, would be given to a good approximation by the expression: | Ip = Ip + T 5A (38) where YT is the time constant of the tank circuit amd equal to KC. This relation is a generalization of an early step in the analysis presented by Shiff and Evans (15). The value of T ranged from about 1/3 min. for the highest counting scale to one min. for the lowest ranges. I, was the true measure of radioactivity; and the activities were computed from its values that were calculated by the above expression from the millismmeter currents, Ip, which varied with time. In analyzing the results, graphs for each experiment were prepared from the data that were collected. In some ~32- cases the concentrations in the effluent stream were plotted against time, in other cases the function, log (< a) } » appearing in Hinshelwood's equation was “drawn, and on oceasions beth curves were prepared fer com- parison purposes. Also, when they were determined, the distribution of the activity residing in the bed, according to the scanning or sectioning method, and the accumulation of activity on the charcoal as a function of time were graphed. It was found that these graphs gave a complete representation of the experimental data in a form from which it could be quickly and conveniently comprehended. There were only a limited nusber of experiments performed, and each represented a considerable expenditure of radioactive materiel. Sinee however with the use of the radioactive indicator technique a very considerable knowledge of the charcoal per- formance could be obtained from each test, a complete set of prints for the experimental graphs has been included in the Appendixes of this thesis. Appendix I Graphs for ixperiments with Arsine Containing Radioarsenic Appendix II Graphs for Experiments with Phosgene Containing Appendix III Graphs for Experiments with Phosgene Containing RadiochLorine Appendix IV Graphs for Temperature Measurements in Phosgene Tube Tests The graphs for each test have been grouped together and arranged in the order of experiment number. =33— EXPERIMENTS WITH RADIOARSINE Preparation and Separation of Radioarsenic and Synthesis of Arsine. as'4 was prepared in the 60 in. eyelotron of the University of California at Berkeley by the bombardment of GeO, with 16 MEV. deuterons. It resulted from a den reaction with Ge’?. Two products of such bombardment, in the form of finely powdered germanium dioxide bearing the arsenic activity, were received from Berkeley, It was first necessary to separate the arsenic activity from the target material and then to synthesize arsine from this unstable arsenic, Because of the striking similarity of many corresponding germanium and arsenic compounds the separation of the elements presented some difficulties; however the following procedure proved satisfactory. The target material, about one gram of GeOz, was placed in a 75 ml. long-necked distilling flask, filled with an inlet tube through a ground glass stopper in the top. 50 mg. of arsenic for carrier in the form of sodium arsenate and 25 ml. of aqua regia were added to the flask and the mixture distilled to dryness. Another 25 ml. of aqua regia was added and the distillation repeated to assure the removal of all the germanium. The distillate contained all the germanium, which was then recovered, but none of the arsenic. In aqua regia arsenic remains in the pentavalent state; and as such it does not, like tri-~- valent arsenic, form volatile halides. The residue in the flask was then treated with 20 ml. of 12N. HCl] and 2 ml. of 42% HBr and the re~- sulting mixture distilled, the distillate being collected in a 100 ml. volumetric flask cooled in ice water. A further 20 ml. portion of HCl was added to the flask and the distillation repeated. The distillate contained nearly all the arsenic as AsCl, and AsBrs, the recovery amounting to over 90%. The HBr served to reduce the arsenate to tri- valent arsenic. A small aliquot portion of each sample was precipitated with HgS as AsSg. From the activities of these precipitates the original activities of the samples were estimated to be 3 and 10 millicuries respectively. A decay curve of the sulphide precipitate, followed for two weeks, indicated that the samples consisted mainly, if not entirely, of the As’4 isotope with a half life of 17 days. The HCl solution of radioarsenic of about 50 ml. volume was mixed with about 50 ml. of 16N,H2S0, containing 0.5 gm. of CdSO,4, a catalyst. The solution was allowed to drop through a funnel onte a large excess of granulated gine (10-12 mesh) in a flask equipped with a side delivery tube. The trivalent arsenic was reduced by the zine to form AsHg, which was carried out of the reaction vessel with the current of hydrogen formed simltaneously. The gases passed through a drying tube and into a liquid air trap where the AsHg con- densed. The efficiency of this synthesis was better than 95%. About 2 liters (STP) of inactive arsine was prepared by the same reaction using larger quantities of zinc and a 10N, H250, solution saturated with AseOg and containing a little CdSO,. With the higher concen~ trations of arsenic only about a 50% yield was achieved. The inactive areine was condensed in the same trap with the radioactive material. After evacuating the trap at liquid air temperatures, the arsine was allowed to expand at room temperature into the glass bulb from which the steel cylinder was filled. Analyses. In the cylinder the concentration of arsine slowly decreased, about 5% per day, the loss being due to the slow oxidation of arsine by oxygen. Even flasks of arsine at atmospheric pressure containing a little air showed gradually increasing deposits of arsenic mirrors on the walls. Because of this decrease in concentration analyses of the air-arsine mixtures were made at frequent intervals. The method of analysis, recommended by the Chemical Warfare Service at the time, proved to be unsatisfactory. According to directions the air-arsine mixture was passed through two absorption bottles containing standard iedine solution and a third bottle with standard sodium thiosulfate solution to retain the iodine vapor. Titration of the contents of the three bottles with either standard iodine or thiosulfate followed. But the radioactive tracer method indicated that the absorption of AsHg was incomplete even at prohibitatively slow flow rates. The method of analysis finally adopted was still dependent upon the action of iodine according to the reaction hHgO + hE, + AsHg =12I7 + HgAsO, + SH’. A 500 ml. flask of measured volume was equipped with a short stoppered burette and openings through which a gas sample could be drawn and about 20 ml. of a NaHCO, and fresh starch solution admitted. Standard iodine was added from the burette to the bicarbonate-starch solution until the blue starch iodine color persisted even after vigorous shaking to assure absorption of the arsine. Unless specifically stated otherwise, all concentrations, volumetric flow rates, and rate constants reported in this thesis have been calculated for the temper~ ature and pressure conditions of the laboratory, consistent with their use in the theoretical equations. Two radioactive arsine~air mixtures, SA~-2 and 343, were pre- pared from the first, 3 millicurie sample, and one, SA-4, from the 10 millicurie sample. The initial arsine concentrations, ¢,, and the initial activity, I,, for these concentrations in the pot under the counter have been given in Table I for the three mixtures. Table I GONCENTRATIONS AND ACTIVITILS FOR ARSINE- AimdArsine ; c, (initial) : I, (initial) Mixture : mg. /lie : counte/min. : t SA~2 $ 7.0 : 44300 SA~3 : 6.2 : 2,900 SA—k, ; 5.6 : 19 ,600 Tube Tests. Several tube tests were performed with each of the three gas mixtures. CWSNL9TUS was used exclusively together with dry air~arsine mixtures (4 2%RH). All tests were made at room temper-~ ature. The concentration of the inlet stream taken directly from the cylinder varied from test to test because of the gradual oxidation of the arsine in the tank. ‘The gas stream from the pot under the counter was passed through a test tube containing dilute aqueous AgNO, solution. The presence of arsine was indicated in the solution by the formation ef a colloidal precipitate of metallic silver, according to the reaction: AsHg + oAg* * 3H.0 = 6Ag * HgAsO, + 6H*. -37- Sueh a chemical test provided an integrated indication of arsine concentration over the period that gas bubbled through the solution, Wath the radioactivity measurements it was possible to detect 0,02 to 0.03 mg./L. of AsHs. The break times for these concentrations were recorded. It is apparent that the definition of break time must include the concentration of gas detected or an equivalent statement. The conditions of the various tests, together with the silver nitrate and the radioactivity break times have been collected in Table II. Bed depths ranging from 2.5 to 5.0 cm. were tested; the flow rates were about 9.1 mL/om’ sec., except in test SA2-4 when a flow of 4.6 mL/em®sec.was used, ‘The silver nitrate break times were considerably shorter than those determined by radioactivity. The gas mixture, SAL, possessing a higher specific activity than the other mixtures, permitted the determination of smaller concentrations of arsenic. In the SAK series, accordingly, it was found that the silver nitrate break occurred with arsine concentrations of about 0.004 mg./L. Table II illustrates very clearly that break times were not reproducible with any degree of precision, especially when the bed depth of the absorbent was 3 cm. or less. The graphical representation of the radioactivity measurements has been included in Appendix I. The concentrations given on the graphs for arsine experiments were all caleulated for (STP) conditions. To test the validity of Hinshelwood's equation, number lia, the values of log ( <2 ~ 1) for each test were plotted against time. In no case were these curves straight; but for most cases appreciable portions of the graphs were moderately straight. The values of k, for Hinshel- ~38— Table II ARSINE TUBE TESTS. CWSN19TUS (Dried) Expt. : Flow : Sg ¢ Bed Depth : res Time : Break Time: ky NO. i > Influent * : (at 0.02 to) : AgNOs 3 ‘ SOC. & Cone. : : min $ 3 mols : : : ° > @ine 7 , mg/L. , : : : % $ 3 $ 3 3 SA2—-2 ; 9.12 8 7.0 3 §.0 3 52 : 38 + 1065 H a 3 $ 3 ‘ SA3-L ; 9.15 3 6.2 : 4.7 t 71 2 «AY : 12100 : Hy t : : g SAkm~2 903k : 5.6 t 5.0 : 49 > BA : 1590 8 t 3 : i : : 3 : ‘ 2 3 Shhwk +: 963k : kel : 4.0 3 43 : 26.7 : 1520 2 : t : : : : 3 i : 3 : SAb-3 2 9.3K : eT t 3.0 t 2he2 : de > §©28h0 ¢ a 3 g 3 SA2-5 t 8.9K 3 6.8 3 2.5 : 8 : 2 : = 1700 3 $ rd : $ H SA2~-6 =: 9,20 $ 68 25 3 12 > 5 + 2100 $ : : H : t wood's equation were calculated from the slope of nearly linear portions of the curves, and they also have been included in Table Ii. ‘The values so caicuisted were surprisingly constant for uniform flow rates and variable bed depths from § down to 3 cm.; when the bed depth reached 3 cm. or less, k, increased rapidly. An increase in ¢, seemed to cause a small decrease of k,. The value of k) for experiment SA2-4, with about one half the flow rate of the other tests, was only one third the value for the comparable tests. These variations of k, were not predicted by the simple kinetics ~39~ considerations of Case 1. It is to be noted that the curvature of the log ( a - i} curves was always positive; therefore the plot of the function, log 2 , which appeared in equation 3h, would be less satis- factory. If k,, calculated from effluent concentrations of tube tests, is to furnish a satisfactory measure of the effectiveness of a charcoal, a@ correlation of these variations will be necessary. When air was passed through charcoal which had been run even well past its break point, no active arsenic could be detected in the effluent In the S3A4 series the 3 mm. slit arrangement was set up to measure the accumulation of activity in the uppermost layer of the bed as a fumetion of time during the ab- sorption of arsine. At the top layer the concentration of the gas stream was constant and equal to ¢,; therefore the complication of variation in ¢ was avoided. Thus the dependence of the rate of arsenic pick-up on N, the concentration of absorbed arsenic in the charcoal, was determined to test directly the validity of kinetics equations 3 and 25a-b, which were assumed in Case 1 and Case 2 of the mathematical treatment. Since the correction for scattering for this type of experiment was somewhat uncertain, both the corrected and uncorrected activities were plotted in the graphs. It was apparent from the curves that the top layer of charcoe] had not reached the saturation of its effectiveness in removing arsine, even for a long time after the break point; but the rate of removal did decrease rapidly with time in the initial stages of the tests. Thus the results were con- trary to the assumptions involved in both Case 1 and Case 2. ~~ In Case 1 equation 15 would give for the activity at x = 0 Activity = K,(1- “kicot ) » Ky, @ constant. (39) According to ecuations 29 and 43 of Case 2 Aetivity =Kat for t <ty, Ke a constant. (40a) and Activity = Ket, for t > t, (40d) As a matter of fact the experimental curves were represented very closely by an equation of the form Activity = ks (1 - aket ) + ket (41) where kg, ky, and Ke are constants. Thus when the first term in equation 4] had reached a nearly constant value, after 40 to 60 minutes of gas flow, the top layer still continued to remove arsine at a constant rate for the remainder of the tests periods, A theoretical treatment, based on this equation, might lead to interesting results; at present it has nob been attempted, It may be pointed out that the value of ky calculated from the slope ef the Log ($2 - i) curves and one point from the curve might be combined in Hinshelwood's equation to yield a value for N,, the saturation capacity of the charcoal. In light of equation 41, however, such a calculation would have little significance. When air was passed through the charcoal after the alr-arsine mixture, no detectable decrease in the activity of the top layer occurred. Because ef the absence of any observed desorption and because a considerable evolu- tion of heat accompanied the removal process, it was believed that the arsenic was completely oxidized on the charcoal surface to form nonvolatile arsenic or oxides of arsenic. After most of the tube tests the distribution of activity in the bed was determined. In experiments SA3-1 and 3~2 both the sectioning and the slit scanning techniques were employed. When the results of the sectioning were corrected to give the same numerical activity for the top 0.5 cm. inerement of the bed es the scanning, identical curves were obtained for the two precedures. For experiment 3-3 only the sectioning technique was used while the more convenient scanning method was employed in the remaining experiments. In every case the scanning activities were corrected for scattering according to equation 37, discussed previously. The points in all experiments fell closely along smooth curves that were drawn to represent the distribution functions. in test SA2-3 the gas flow was stopped at approximately one half the break time. In this bed no activity was detected in the bottom layer, and an estimated 25% of the activity remained in the first two om. of the 5 cm, bed, Thus the lower 3 om. constituted an unused or "dead" layer. With beds for which gas flow had proceeded well past the break, the bottom layer frequently possessed an appreciable quantity of arsenic activity. The direct com- parison of the distribution curves with the theoretical equation 15 of Case 1 and equations 29, 41, and 43 of Case 2 for N did not appear practicable, However a useful correlation was suggested from the results of Case 1. Equations l4b and 15 were combined to give nC = hdd + / (42) and wos’) (2) (43) if N, is eliminated between these equations, there obtains no (2 *) a “Aye ge’ “Tee (44) e + i When t’ becomes sufficiently large, the following simple expression results. ! N(O,b" = o(Lit’) (45) Sy Bat N(L,t / N(O,t’) would be given by the ratio of the activities on the charcoal in the bottom and the top sections; and e(Lyt )/¢g would be the final effluent concentration divided by the influent concentration, The experimental values for these quantities appear in Table III. Table III DISTRIBUTION OF ACTIVE ARSENIC ® @ L t % we | 4 ' t No, ? t : * : A “ aa ; 9(,t) * Bed Depth * Time of Gas * Break Time ' 2s 3 So ‘ ; Flow * (0.02 to 0,03)* ; ; em. ; * (mg./L. Ass )* ‘ : ; Min. : : : g 3 a z Min. 3 t SA3-3 340 3 93 : 37.5 : 0.4 : 0.40 SA-3; 300 80 she 0.5 : 0.36 SA3-2 het 122 : 52.5 : 0.5 : 0.35 SAb~k = $ oO 3 214 : 43 : 0.6 ¢* 0,80 SA2-3 5,00? 17.5 * ea. 35 ; 0.0 § 0.003 SAQ-1 * 5.0 ! 72 ° 52 ; 0.1 * 0.05 SAme2 > 5.0 110 49 0.2 |; 0.20 SAML , 5.0 | 170 : 71 : 0.6 ; 0.59 The correlation predicted by eouation 45 seemed remarkably good in many instances. in the bed could be calculated from the data. Although N, could not be obtained, the concentration of arsenic Such a ¢alculation was made for experiment SA4-~4,, the longest run, from the known arsine in- put and from the loss of arsenic determined by a graphical integration of the effluent concentrations for the test. Then with the aid of the experimental distribution function it was calculated that the concentration Whe of arsenic, expressed as the element, in the top layer of the charcoal amounted to 118 mg./cc.’ The data accumulated with radioarsenic. have been by necessity very limited. For this reason it appears unwise at present to draw further conclusions. The experimental results and the conclusions which were reach@i,however, have in themselves proved useful in partially de- seribing the phenomenon of arsine removal. Limitations in the correla~ tions of charcoal performance by present methods hawe appeared; and certainly ‘the desirability of a more satisfying theoretical treatment to predict the experimental behavior has been clearly shown, It is very probable that a more extensive investigation, especially if it included the radioactive tracer methods, would lead to a clearer understanding of the mechanisms involved. =) 5~ EXPERIMENTS WITH RADIOPHOSGENE (C*, C1#*) Beeause of the short half lives of radiocarbon and radiochlorine the work with phosgene was performed mostly in the laboratories of the University of California at Berkeley so that experiments with the radio~ active materials could proceed immediately following the cyclotron bombardments. Preparation of Inactive Phosgene. About 400 g. of phosgene was prepared from CO and Cl, by passing an approximately equimolar mixture of the dried gases through two beds, each about 5 em. leng, of dried GWSNL9, which acted as a catalyst. Cylinder Cl, was used, and CO was generated by the reaction of concentrated H,50, with formic acid. The produet gases were passed through a bed of crushed antimony which com- bined with the free chiorine and condensed in a flask cooled in a dry ice bath. The phosgene was further purified by refluxing the liquid in the presence of metallic antimony for a period of seyeral hours, the reflux condenser being cooled by a dry ice bath. Ali but a trace of yellow color due to dissolved Cl, was removed. The liquid phosgene was stored in a emall steel cylinder . gil Preparation of Phosgene Containing Active Carbon. was pre- pared by a den reaction from Bo, the target, in the form of boric oxide fused on a copper plate, was bombarded with 80-100 4a. of 16 MEV. deuterons for 30 to 60 min. in the 60 in, Berkeley cyclotron. Active carbon prepared in this way existed largely as CO. The target was en- closed in a gas tight brass box with two ports. To each port was attached a 2 L. bulb; one bulb was evacuated and one contained dry oh Qe COQ at one atm. Shortly after bombardment the radioactive CO was flushed into the evacuated bulb with CO from the other bulb. ‘The previously evacuated bulb,containing CO at 0.5 atm. and bearing all the active co, was filled to one atm. with dry Clg. Illumination of this mixture for 10 min. with light from two 500 w. incandescent lamps placed very close to the glass bulb, resulted in the synthesis of phosgene with a yield of 80-90%. (This photochemical reaction was the method by which Davy (16), using sunlight, first prepared phosgene). The phosgene from the reaction bulb was passed over metallic antimony to remove Cl, and was condensed by liquid air; the trap was then evacuated to remove any CO. The resulting purified phosgene was distilled from the trap directly inte the evacuated steel cylinder, dry air being used to flush out the trap. When desired, inactive phosgene was also introduced before compressed air from another cylinder was added. The synthesis required about 40 min,, or two carbon half lives, after the cessation of bombardment. The total activity of the carbon, calculated for the end of the bombardment, amounted to about 100 millicuries., Preparation of Phosgene from Radioactive Chlorine. 138, resulting from a d-p reaction with chloride, was formed by the direct deuteron bombardment of LiCl, fused on a copper target plate, with 16 MEV. deuterons at 80 to 100 Ya, for about one hour, Shortly after bombardment the target plate was removed from the cyclotron and the LiCl washed from the plate with distilled water into a distilling flask. The active chloride was converted into chlorine by direct exchange, making use of the fortunate circumstance that this exchange is practically instantaneous (17, 18). Te accomplish the exchange, one L. of Cl, at 1 atm. was bubbled through the LiCl solution, passed through a Mg(C10,). drying agent, and then condensed with liquid air. The Cl, was distilled from the trap into a 2 L. bulb that contained dry CO at 0.5 atm. ‘The chlorine, which passed through the LIC] solution, removed at least 90% of the chlorine activity. The photochemical preparation of phosgene from the CO-Cl. mixtures was carried out as with the radiocarbon. The synthesis required about 50 min. or slightly more than one chlorine half life. The radiochlorine samples were intensely actives and in a con- densed phase, they glowed brightly in the dark. The measured total chlorine activity at the end of bombardment was about 100 millicuries,. Chemical Analysis and Tests for Phosgene. The concentrations of phosgene in the cylinders were determined by bubbling the mixture at a known rate, about 250 ml./min, as indicated by a flowmeter, through two absorption bottles connected in series and each containing 50 ml. of 3. NaOH solution. The chloride content of the NaQH solutions was then determined by the Volhard method. A saturated aqueous solution of aniline and N,N’ ~ diphenylurea was used for the chemical detection of phosgene. The formation of the white precipitate of diphenylurea (solubility in H,0 0.015 g./100 ml.H,0 at 25° ¢.) gave a less sensitive but more reproducible test than did p~dimethylaminobenzaldehyde-diphenylamine paper or solution, which were used occasionally. The sensitivity of the aniline test was estimated by passing known concentrations of phesgene through the solution and noting the time required for the formation of a visible precipitate. This was made the basis of a rough chemical estimation of minute concentrations of phosgene. Subsequent experiments have not given quantitative reproducible results, and the concentrations obtained by this method were considered to be lower limits of the phosgene concentration. The method was useful however since it indicated only phosgene while the radioactive method re- corded total chloride concentration. The sensitivity of the test as used in the experiments varied from 0.15 to 0.40 mg. of phosgene, depending on the concentration present. These results were obtained under the following conditions: flow rate 2 L./min., room temperature 18.5° C., 5 ml. of aqueous solution saturated with aniline and diphenylurea, Acueous AgNO, solutions proved useful indicators also since with the concentrations in- volved a precipitate of AgCl was formed only slowly when phesgene-air mixtures were passed through them. On the other hand, low concentrations of HCl produced an immediate copious precipitate. The presence of phos- gene was frequently detected by smell. In the first tube test with phosgene con- taining labeled carbon it was evident that large amounts of a substance containing the radiocarbon were transmitted almost without delay. To further study this rapid penetration by some carbon containing material, several, experiments were performed at reduced flow rates (about 0.5 and “1.0 ml./em? sec.) utilizing a 5.0 om, bed of CWSNIQTUS (Dried), (50 RH), and (70 RH) at room temperature. The condi tions of these tests have been included in-Table IVA. Desorption was studied in experiments 6G5—1¢0 and CGg-20 by passing an air stream through the bed after the air- phosgene mixture. In experiment CG4-C the flow rate was increased from Table IV A CONDITIONS OF RADIOCARBON TESTS AT LOW FLOW RATES » 3 : , ot : t Flow ¢ De 4 : ; Ge ; Phosgene , ml./emssec.? : Sorption $ : > omg./L. ? ' Ming + Min. 3 $ : 3 3 3 : : t is.) 2 : : $ i : ( mass H 2 H : $ ;\ = : 3 3 8 $ Hy 3 3 ; 3 CG2-C =: GWSNL9TUS : 20.5 * 9.22 3 0.54 : 65 : : (Dried) : t (93.3) 3 t ‘ 3 & & 8 Hy H CG5—=26 : CNSNL9TUB : 20.6 : 9.62 3 0.53 5 3h $ > (Dried) : > (97.4) ¢ : : t : a. t i t CGb—1C : GWSNI9QTUS : 20.0 ¢% 10.6 : 1.06 : 30 : | : (Dried) = 2 (LO7.1)¢ : : 3 Ps ‘ $ : 3 CG6-20 : GUSNL9TUS : 20.0 * 10.6 : 1.06 4 35 = hh bo 62 : (Dried) : * (107.2) ¢ : ; H : 3 t : : CG5-LC + GWSNLQTUS * 20.6 * 9.62 + 0,55 : 7o $+ 82 to 102 $ (50 RH) 2 > (97.4) = 2 3 : a a & * 3 Coh-c : CUSNLOTUS * 19.8 * 6.94 * (0.53 7 (nm 2 (70 RH) § * (7.1) F ¢ - : $ : : + (10.0 ; (75 to ' t 3 H : : 10 : ? t t : t : Table IV B RESULTS OF RADIOCARBON TESTS AT LOW FLOW RATES (Bed Depth = 5 cm.) Expt. | Total Phosgene, 0" : cf Remarks No. : Input Absorbed Desorbed ? : Hs § : mnole. : mnols. : mnols. : : $ t : ‘ : : : cG2c t t : 2 : : ; ; of activity in entrant : t : ; layer. Oa5-2c ; 03h ; ORT : : 2 : : : : ; completely removed : ' : ; by soda lime. a 2 : t : CG6—26 , 0.78 : 0.29 : OnLy ; Activity in effluent ‘ : : , completely removed 2 ry 2 zg by soda Lime. casmic 5) Q473 * 0.60 oe | : : : 2 t : : : t ( hwd HH ( 0.5 3 ; (Additional) 0.51 to 10,0 ml./em.” min, after 71 min, and maintained at that value for 30 min. The penetration of the bed by active carbon has been illus- trated in the figures for experiments: CG2-C, CGA-C, CG5-1C, CG5+2¢, CG6-LC, and CG6-2C, which are contained in Appendix II. ‘The figures show the carbon concentrations that appeared in the effluent streams. None of the experiments were run sufficiently long to obtain a phosgene break as indicated by the aniline test. The amounts of active carbon retained and transmitted by the bed as well as the quantities later de- sorbed were calculated by graphical integration of the effluent con- centration curves. These results have been listed in Table IVB. The results from the low flow rate experiments were noted: l. The carbon activity appearing in the effluent was partially removed by a single liquid air trap and completely by a soda lime ab- sorption tube. It possessed no detectable odor and later experiments proved that mo chlerine was associated with it. Upon these observations and the nature of its origin it was concluded that the active substance was probably CO,. 2. For low flows the CO, did not appear immediately in the effluent, there was a finite CO, break time. 3. For comparable conditions the presence of water in the char- coal markedly increased the retention of CO,. The CO, break times in- creased and the slope the effluent CO, concentration-time curves de- creased with increasing water content. 4. The charcoal retained from 0.3 to 1.0 millimols of the active carbon; and of this, a sizable fraction, 0.2 to 0.5, was readily desorbed by an air stream. Tube tests were then performed at normal flow rates (ca. 10 ml./cem."sec,) with dry air-phosgene mixtures and with 5 cm, beds of the | GWSNL9TUS (Dried), (50 RH), (70 RH) and GWSN19 (Dried). ‘These ex- | periments were all. run well past the phosgene break indicated by the chemical indieaters. In some instances desorption with an air flow at the same rate as for the toxic stream was studied. The conditions of the tests have been presented in Table VA, The graphs of the radioactivity results for these experiments also appear in Appendix II (Experiments CG7=C, CGG=C, CGIC, CGLO-C, CGLI~C, CG12~C and CGl4-C). The figures show the plots of the active carbon concentration in the effluent. ‘the total phosgene input and the chemically indicated break time for each experiment have been included in Table VB together with the amount of carbon retained during absorption and the amount later desorbed, which were obtained by the graphical integration of the graphs. The following results were noted: — 1. A substance containing radiocarbon, completely removed by soda lime, penetrated the bed from the start of gas flow; it was con- cluded that this substance was again CO,. 2. The concentration of COg inthe effluent in a very few minutes attained a constant value which was equal to that for the influent radio+ carbon, The total carbon concentration did not change later in the ex- periment when phesgene penetrated the bed. 3. Unexplained discrepancies appeared in the results of experi-~ ments CG10-G and CG12-C which were performed under almost identical Table V A CONDITIONS OF RADIOCARBON TESTS AT HIGH FLOW RATES (Bed Depth = 5.0 om.) Expt. * Charcoal * Temp. * fe ; Flow Rate ; Duration * Duration No. ? ‘ a ; + Flow ¢? Desorption : s °C. 3 Phosgene * mi. /em@sec.* fk ; ; * ng/L. ° * Min. ; Min. 3 : t ( seghes| $ : : $ : ; $ : : CG7—C ; GCNSNLOTUS ; 18 : 9,00 3 10.4 3 80 : 110 to 118 : (Dried) ; : (91.0) ; : : cos-c ; GUSNLOTUS : 18 : .h7 : WO. : 105 +: 112 to 122 ; (Dried) ; : (75.6) : 3 3 CGlL~C ; GWSNLOTUS ; 17.5 : 7:86 : 103 : 90 + ULtolTl : (Dried) ;: : (79.6) : t : ca9~c ; GWSNIOTUS : 16 : 890 : 102 : 100 +: 120 to 12 : (7O RH): : (90.0) : : : cGlo-c : GWSNI9 «i: «18 tCO708H tC ct OS ~ : (Dried) : : (79.6) 3 : : 3 $ 3 $ $ M4 CGl2—-G + GVSNL9 : 16 3 7.51 3 10.3 8 90 8 109 to 127 : (Dried) : : (76.0) 3 : : CGlAeG ¢ CISNIOTUS + +25 : 2202 3 8.0 3 50 3 (69 to 86 : (7O RH): : (22k) 3 3 : ¢ : ¢ $ 2 t 2 §6(88 to 138) 6 3 i 3 8 $ (at 19°C, ) ~Sijo Table V B RESULTS OF RADIOCARBON TESTS AT HIGH FLOW RATES (Bed Depth = 5 om.) % cy cooled to -25°C. Aniline break 20 min. No phos- gene detected in de~- sorption. Bed warmed to rm, temp. during second desn. Expt. ‘Phosgene ° Cc * 6 * Phosgene ' Remarks No, | Input ' Absorbed + Desorbed ' Desorbed | : mnols. ' mmols. * mmols. * mols. : t 2 t H t t t t t CG7—“C : 149 : 0.5 tcea.0.1 i : C* completely removed : t t 2 ¢ from effluent by soda : : : : t lime. Ca. 3 min. after $ t : 2 t desn. started, Phosgene t t 3 8 ¢ gone, was 7 x the C* conc. : t 2 2 t CGB-C : 15.7 3: O45 : O.17 3 : GC completely removed ? : : $ $ from effluent by soda 3 H 2 $ : ime. Continued strong 3 $ t : ? phosgene test by aniline t 3 : t : after initial burst of : t : : + = =6activity during desn. CGll~C : yeh ¢ O64 s 0.13 = 60.55 $ Aniline break, 60 min. CG9-C os: 17.8 $s 1.2 : 0.09 ‘ca.0.05 * Large desorption of HCl t : t t t CGLOe-C 2 LOsk 3 & : : AgCl break 18 min. : : $ 2 * Aniline break 50 min. : t 3 : f CGl2—-C : 13.7 ¢ O.2 5 0625 & OL 7K * AgCl break, 10 min. : t : 3 * Aniline break, 57 min. CGly-C : 20.5 ¢* 1.42 $ 0.2 = 0.05 * Bed and influent stream : : : t oo $2 of SF Ss £9 OO conditions with CWSN19 (Freshly dried). In CG1O-G the active carbon in the effluent stream rose to only 70-80% of that in the influent. A& the aniline break the active carbon concentration may have started to rise, although the test was not continued long enough for this to have been determined with certainty. In (G12-6, however, the effluent concentration rose quickly to the influent value and remained con- stant thereafter. HCl appeared in each test in the effluent in 10 to 20 min.; but phosgene did not appear until 50 to 60 min., the phosgene break time being somewhat shorter than for the dry whetlerite. 4. The quantity of active carbon retained by the whetlerite and and the quantity that could later be desorbed were of the same order of magnitude as the corresponding values at the low flow rates when the total phosgene input was approximately only 5% as great. 5. A surprising result was that in the desorption experiments with dried CWSN19TU8, phosgene continued to be desorbed, as indicated by the estimates with the chemical tests, after the carbon activity had fallen to a negligible amount. A lower limit to the quantity of phosgene desorbed was given by the precipitation of diphenylurea from aniline solutions. These results indicated that three to four times the amount of phosgene was desorbed than was te be inferred from the radioactivity. Experiments with Radiochlorine. The preceding experiments, although informative, proved that radiocarbon could not be used successfully to evaluate the protection of the charcoals against phosgene. Accordingly, phosgene labeled with e138 was prepared and used for tube tests with dried and equilibrated whetlerites in a 5.0 cm. bed at room temperature, ~25°C, and 40°C. Again dry air-phosgene mixtures ( (2% RH) were used exclusively. The radiochlorine, in contrast to the active carbon, ex- hibited normal behavior. The conditions for the tests, together with the aniline break points and the 1% chlorine penetration times, have been listed in Table VI A. The graphs for the experiments, CG16é~-Cl, CGL7~C1l, CG18-201, CG20-1, 2, 3C1, CG2l-1, 2, 3C1 and CG22-Cl, have been included in Appendix III. Active chlorine concentration vs. time curves were prepared for comparison with the radiocarbon tests. Also the Log/ = ~ 1) function was plotted to test the theoretical equations. In the few cases where it was measured, CG17-Cl, CG18-2C1, and CG22-(1, the accumulation of activity in the entrant layer was plotted. On the graphs the aniline break points have been indicated by the symbol, A. For determining the accumulation of activity, the slit technique was employed. Alor 3 mm. slit was used; and because of the in- tensity of the radiation the slits were 200 mm. deep, the top of the slit being in line with the top of the bed. The correction for scattering was not determined with radiochlorine; therefore the activities were plotted uncorrected. The correction for scattering should not have changed the general form of the curves however. It appeared that the concentration of chlorine on the entrant layer of charcoal during absorption did increase with time with a rapidly de- creasing rate. “It did seem to approach a limiting value which was nearly reached at the break time (effluent chlorine = 1% influent). Thus it agreed fairly well with the equation resulting from Case 1 5 Ju Table VI A HADIOCHLORINE TESTS (Bed Depth = 5 cm.) Expt. ? Charcoal ¢ oo * Flow Rate * Break Time * Break Time Bo. : ; Phosgene . ml. /em. sec. : Cl* ig of : Aniline (Temp. ) ng. /L. Duration : Influent ‘ min, : , (uumels.CL®) Flow : min, : cas—cl ‘cwsnigrus £ 7.82 | 10.3 ; 60 62 (19°c.) ? (Dried) | (151.8) | (200) colé-cl jcwsmugrus { 9.38 ; 10.2 { bh : A (19°C.) , (Dried) , (189.8) . (80) : : ca22-ch fowsnigrus | 22.3 {| 10.3 . 6 re (2u%c.) | (Dried) , (451) ; (71.8) : : ClS-201 icusMigTUe | 7.36; = 10.3 , ke ee. (20°¢,) , (50 RH), (149) : (95) : ; CG17-Cl jGWSNI9TUS ; 10.6 4 103 ; OT os (19°e.) } (70 RH) =; (22a) ; (210) : : 0G20-3C1 ;GHSNL9TUS ; 23.9 3 8.0 Os ; 80 6G20-201 fcHSNI9TUg , 23.9 + B60 : 25 : 39 (=25°C.) , (50 RH), (484) : (60) : 3 ES = 3 3 $ CG20-1C1 ;CWSNI9TUS , 23.9 : «8.0 ; 45 res CG21-3Cl ,CWSNLOTUS , 20-4 |; 10.0 (5 > £37 (40°C. ) 8 (Dried) ¢ (413) 2 (58) 3 3 CG21-1Cl .CWSNI9TUS ; 20.4 ; 10.0 ;R : <33 (40°C.) , (50 RH) =; (413) : (59) : 3 CG21-2CL .CWSNI9TUS , 20.4 : 10.0 : 28 > <35 Table Vi B RSSULTS OF RADIO CHLORINE TESTS (Bed Depth = 5 em.) Expt. = Phosgene * Total cu" cl” * Fraction * Remarks No. : Retained * Absorbed ° Desorbed : oy ° * Before : ; ned ° * Break * mols. * Time Interval * Desorbed + ; ° * of Desorption * ; : mnols, : : . : & 2 3 (min. ) o 6 CGl5-Cl 2: 9.57 3 3 g t $ $ g g : CGl6-Cl : 8.39 3: 29.2 t 3.80 : O12 : Desorbed first : ¢ ¢ (135 te 16h): 3 with Ne, then 2 2 : (165 te 186): : with air. Aniline 2 t z : : test + during 3 g : z : desorp. Fs 2 3 3 $ CGe2—-CL 3: 11.83 : 41,0 : 17.0 2 0.33 $ Aniline test ++ 3 3 : (7h to 130) 3 : during desorp. CGLS-2C0L : 6.31 : 26.8 : LAL : 0.05 : Aniline test faint 2 3 : (110 to 139): ¢ AglOg test + : $ : 8 : during desn. CGL7-CL 3: 9¥.61 : &1e§ z heO5 s 0.10 : Aniline test very # : : (191 to 216) 3 s faint. AgNO, test + 3 t : 3 : during desorp. g 2 : 3 : CG20~-3C1 2 32.5 : 83.3 3 lil 0.13 3 Aniline test ++ : 3 : (111 to 127) =: : during desorp. CG20-201 : 10.97 3 bBeS of 2.62 : 0.54 ¢ Aniline test + : g : {61 to 75) 3 $ AgNO, test ++ : 3 3 3 2 during desorp. $ : 3 3 3 CG20-1CL : 6.60 : 36.2 3 3) 2 0 $ Aniline test : : : (58 to 65) 3 ? negative during : % 3 8 § desorp. CG21-3C1 3: 11.89 ¢ £2.0 H 11.3 = 0.28 3 Amiline test ++ 3 g : (68 to 83) 2 3 during desorp. GG21-1C1 : 12.80 : 3.7 3 5.37 ® 0.13 $ Aniline test ++ 2 : : (53 to 68) 3 * during desorp. CG21-201 3: 13.20 $ * : ~ Sw $ Aniline test + : : 3 3 * during desorp. NOTE + amd ++ indicate positive and strongly positive test. ~59— ( “Kre0t} Activity = K \l-e » K a constant. (39) The steeper rise for CG22-Cl may have been due to the greater slit width (3 ma. ss compared with lmm.). The dependence of the rate of absorption on N would therefore be in agreement with the assumption for Case 1. The resulting equations for Case 1 then would or would not be valid according to the dependence of the rate of absorption on ¢. It was noteworthy that om third to one half of the chlorine in the entrant layer could be desorbed; the remainder was not removed by continued air flow, From the plots of 10s £2 - } it can be seen that a small fraction of the entrant chlorine appeared in the effluent and remained roughly constent at about 0.1% of the influent until the break. This was noted whether the thermocouples were or were not inserted in the bed. To exclude the possibility that this effect was caused by radia~ tion from the charcoal scattered through openings of the counter shield, a liquid air trap and a seda lime tube were inserted in the influent line before the pot and counter assembly. Each of these removed a large fraction, although not all, of the activity from the stream. It was therefore concluded that some chloride containing substance pene« trated the charceal in small quantities immediately, Further investi- gations would be required for its identification. The radioactivity break times for 1% penetration of chlorine proved to be nearly identical with the aniline breaks. The values obtained by both methods have been given in Table VI A. The break times for comparable tests were not strongly dependent on the moisture content of the charcoal at either 20 or 40°C, although possibly the ~60= higher water content contributed to slightly better protection. No difference in protection between 20 and 40°C, was apparent. The differences in moisture content however resulted in striking differences of protection at low temperatures. At -25°C. the dry charcoal had three times the life, the (50 RH) sample nearly the same life, and the (70 RH) sample roughly only 0.6 the life of the absorbents tested at room temperature. The quantities of phosgene retained by the charcoal before the break together with the amount of active chlorine absorbed by and desorbed from the bed, as calculated by means of graphical integrations of the effluent concentration curves, have been given in Table VIB. The molecular state of the chlorine was not indicated, of course, by the radioactivity measurements. However a comparison of the aniline and silver nitrate indicator tests showed that with dried whetlerite only very minute braces of HCL compared to phosgene could have been present in the effluent during either gas flow or desorption. The concen~ trations of chlorine during desorption were of the same order of mag- nitude as the concentrations penetrating the bed in the later period of absorption, and usually they decreased fairly rapidly as desorption progressed, The value of 3.80 mmols. of active Cl desorbed as phos- gene for test CG16-Cl can be compared with the value, 0.13 mmols. of active carbon, for experiment CGl1-C. The dried whetlerite from one of the tube tests still retained the oder of phosgene several days after the test. When it was placed in water, both cupric and chloride ions were identified in the solution. With the equilibrated charcoals at 20 and 40°C., on the other hand, although the chlorine concentration curves were similar to those for the dry whetlerite, it was evident from the chemical indicators that the effluent contained both phos- gene and HCl. HCl, and CO, are, of course, the hydrolysis products of phosgene. During the last periods of gas flow and during the desorption probably the larger fraction of the chlorine was present as HCl. In general, there was less chlorine desorbed from equilibrated chars. At -25°C. a surprising result was that no detectable chlorine was de- sorbed from the (70 RH) whetlerite either at the low temperature or after the charcoal had been warmed to room temperature, although the break time had been only 15 min., and HCl had appeared in the effluent. It was found from the log ee -~ 1) graphs that, as was the case with arsine, straight lines were not obtained for this function, but that appreciable portions of the curves were nearly linear. By means of the approximate Hinshelwood equation, ¢ ky NAM fae 1a( <2. 1) = -kyeot * ae 3 (he) the values for the rate constant, k,, and for the capacity factor, Ny, were calculated. These parameters were computed for the point at which ¢,/e = 100 since this point corresponded closely to the chemical break point, and because the logarithmic plots had only slight curvature in this region. The calculated values appear in Table VII. Although the number of experiments was too small to permit satis- factory generalizations, some interesting observations concerning the Hinshelwood constants may be made. The value of k, for dried CVSN19TUS at room temperature in the two comparable experiments decreased slightly, 1 from 880 to 730 cm:(mmols. cL)7 min.” with a two fold increase of ¢,. This difference may not have been significant. However the values . wot ow Table VII PARAMETERS OF APPROXIMATE HINSHELHOOD EQUATION, NUMBER 17, CALCULATED FOR RADIOCHLORINE = (3-3) ok. ct 1 k, AN o* y CG17=Cl CG21=301 CG21~LC1 CG20-3C1 ixpt. ° Charcoal * Tem.* vo * 4, ; k y No. ° 6 AF @ 3 a 3 ’ CWSNIOTUS © “Cc. : : 3 : . cm, | mmols, Cl . cm. mmols. Cl : : min. : mm : (mmols. Cl)min. cm? CG16-Cl : (Dried) 19 618 : 189 : 880 1.68 : : : OG22-CL : (Dried) 21 618s AS] t 730 2020 CG1L8-2C1 : (50 RH) 20 618 ss 1s9 3 50 1.34 CG20—201 CG20-3¢C1 ae te oS 6 e& G8 68 Oo WO Bo «Oo oH (70 RH) (Dried) (50 RH) (Dried) (50 RH) (70 RH) eo sO 20 8S SS 5H Ge 62 oS 66 G2 Ge be Ce 60 *0 OF oH ev Co #9 Ge OO Ge of S09 of 66 OF Be C8 O89 Of OS FO OO Of fon # OF oO 8 GO BH 618 ae 80 £% be 413 413 48h, Se ob OO of 370 480 LL0 ee 8 00 Co 6S O88 6 O88 Of Ge G0 8S Ge 26 GO Oo 6H ao 90 fee of SD Be So Ge oe 3.06 1.72 1696 bye 20 2.18 1.70 decreased considerably with increased water content, being 540 for ~1, CWSNL9TUS (50 RH) and 320 cm°(mmols. C1) min. ~~ for (70 RH). The rate constants were markedly higher at 40°C. than at room temperature, amounting to 1150 and 1000 om: (mols. cr) "tmin. for the dried and (50 RH) whetlerite. The capacity figures, Nj» on the other hand, showed a rather large increase with inereasing concentration, increase with increasing water content, and a significant decrease, as would be expected, when the temperature was increased from 20 to 40°C. Thus, the variations of these parameters indicated considerable differences in charcoal behavior under the different conditions tested. However in the tube tests performed, the variations of k, so balanced the changes in N, that for equal concentrations nearly equal break times were ob- tained in most cases at 20 to 40°C, At -25°C. the values of k, were nearly independent of moisture and were about equal to the value for the (50 RH) material at room temperature. The striking differences in the protection occurred at the low temperatures because Ny varied from the low value of 1.70 mmols. Cl/em. with CWSNL9TUS (70 RH) to the extremely high value of 4.20 auole. Cl/em: for the dried whetlerite. The desorption of chlorine which occurs with phosgene was not in agreement with the assumptions involved in Case 1 for the derivation ef Hinshelwood's equation. However it might result thet the overall differential equation, including desorption as well as absorption, might approximate the assumed equation 3 for a constant concentration of toxic material entering the bed. From the limited number of data the use of the results from Case 1 to evaluate a chareoai's phosgene protection shows some promise. The dependence of ky and N, on the test conditions should however be determined, ~bh— The values of k, and Ny were calculated on the basis of chlorine. Vhen the values of ky were changed to a phosgene basis and expressed with the same units as for the case of arsine, very similar values were obtained for a pair of comparable experiments: for arsine, ex~- periment SA4-2, ky = 1590 L. (mols. AsHs ya; for phosgene, ex- periment CGL6-Cl, k, = 1760 L.(mols C0Clg)"4min.~*, Doth these ox- periments were with 5.0 cm. beds of dried CiSNL9TUS and at nearly the same temperature and flow rate. The value of No, expressed on a phos- gene basis for experiment CG16-Cl was 0.84 (sols. OCLs) fem. The maximum concentration of arsenic in the charcoal was found to be 1.57 (mols. AsHs)/om: thermocouple Measurements. Temperatures at the top, in the middle, and at the bottom of the absorbent beds were measurd@ by means of ther~ mocouples whose junctions were inserted in the charcoal 0.5 cm., 2.5 cm, and 4.5 cu. above the bottom of the bed. Curves of the temperatures de- termined by these thermocouples have been included in Appendix IV. At low flow rates the temperature change in the bed was small; but at high flows a temperature rise of several degrees accompanied the removal. The entrant layer showed little temperature change; this was attributed to the removal of heat by the incoming stream. jith this qualification it was seen that a warm zone moved down the bed. ‘The chlorine break occurred close to the time when the peak of the temperature wave reached the bottom (exit) of the charcoal. Quantitative interpretations of the temperature measurements would be difficult because of the heat less to the room; therefore no estimate of AH for the removal process was attempted. The maximum rise in dry beds wes about 6°C.; for moist ~65— beds it was 1-2°C. since probably H,0 was desorbed by the gas stream, In the early tests desorption was carried out with air from a compressed air line; these desorptions were accomp:nied by a rise in temperature. In experiment CG12 and subsequently, a calcium chloride tube was inserted in the line; thereafter no marked temperature effect was mobed. | ion of Results. The removal of phosgene by whetlerite would appear to be more involved than proposed by the present theories of adsorption and of hydrolysis by water present either in the charcoal or in the gas stream (19). Two major features of the experimental re- sults aust be explained; they are: 1. The carbon dioxide transmitted was equivalent to nearly all the phosgene which had entered the dried charcoal up to the time a chloride compound penetrated the bed. 2, Greater amounts of phosgene than corresponded to the labeled carbon were desorbed from a dry bed by an air stream after absorption, The first hypothesis which suggests itself to account for the evolution of CO, from the dried whetlerite is that the copper oxide existing in the charcoal played an important role in reacting with the phosgene, according to the overall reaction cuo + C*OCLe = CHO, + CuCl. It may be pointed out that actually this reaction may have occurred in two steps in which the phosgene was first hydrolyzed and then in which the HCl formed was neutralized by the Cud. ‘Thus, C¥OGLle + HeO = C%Og + 2HCL The water formed in the second step would therefore be released to hydrolyze a second phosgene molecule, and so on. In this way a trace of water, which was not removed at 150° C. or which might have entered the charcoal in handling, could contribute to the hydrolysis of a sig- nificant quantity of phosgene. The maximum possible water content of the air used, 24 RH, would itself have accounted for the hydrolysis of about 2 mg. phosgene /L. according to the first step of the above reaction pair; and with the formation of water in the secend step it would also have been able to remove additional phosgene. If it were assumed that the charceal adsorbed the HCl as such, however, without the formation of copper chloride, then to account for the CO. formation by hydrolysis in experiment CG22-Cl would re~ quire that the charcoal must have contained 5.4% HeO, a value that seemed unreasonably high. It is worthwhile to note that the charcoal was stored in tightly sealed screw cap bottles, and that, further, samples which were redried at 150°C. and used the same day or a few days later gave the same results as the older samples from the storage bottles. The absence of appreciable HCl in the effluent stream at any . time after the Cl™ break or during desorption indicated that if any acid were formed, it must have been retained very firmly by the char~ eoal. Indeed, if the phosgene had hydrolyzed, and the HCl] had not been neutralized, then HC] would have greatly predominated over the phosgene in the bed, since nearly all the phosgene until the break had reacted .to form CO,g. Then HCl, contrary to experiment, would have been ex~ pected to appear in the effluent. In experiment CCG16-Cl it was cal- culated that 16.8 milliequivalents of chlorine had been retained by ~67-~ the bed before the cl” break; assuming the whetlerite to have been 7% copper, the amount of copper present was 17 milliequivalents, that is, just about enough to account for all the phosgene or HCl which might have formed. The behavior of the CISNL9 base char would appear somewhat sur~- prising therefore since with it, also, large quantities of active carbon dioxide appeared immediately in the effluent gases. Because the unimpregnated char contained no copper oxide, it might be sup- posed that the hydrolysis resulting in the formation of COg would alse have produced HC] which would have appeared in the effluent. This was in agreement with the observation that for experiments CGLO-C and CG1l2-¢ HCl was detected in the effluent after 10 te 20 min., about 40 min. sooner than the penetration of phosgene was in- dieated. A 2.5% moisture content would have been required to supply the water for this hydrolysis in CG12-C. The reaction of phosgene with CuO or water does not explain the fact that during desorption more phosgene appeared in the effluent stream than could be inferred from the carbon activity. This was indicated quali- tatively by the chemical indicator solutions; however it can be demon=- strated quantitatively by the comparison of experiment CG11-C with radio- carbon and experiment CG16—Cl with radiochlorine, for which closely similar conditions prevailed. It should be noted however that nitrogen was used for a portion of the desorption in CG16-Cl. According to an integration of the effluent concentration curves for these experi- ments, 0.13 mmols. of c* was desorbed in CGl1-C and 3.80 mols. of Gl” in CGlé-Cl. Thus a ratio ef 15 to 1, based on the phosgene equivalents of chlorine and carbon, existed. This represented a ~bh— lower limit to the actual ratio since probably a portion of the carbon was present as COg, and the desorption in CG16~Cl was stopped while the effluent concentration was greater than 30 maols./:m: It is to be reemphasized that with the dried whetlerites only phosgene, and no HCl, was detected in the effluent stream. A second hypothesis that could account for the evolution of CO, 1s that phosgene combined directly with oxygen on the charceal te produce C"0,, the two chlorine atoms being left momentarily ad~ sorbed on the charcoal. The chlorine might then have reacted with "surface oxides" present on the charcoal to form inactive phosgene. CeOCL2, = C#OCleg (adsorbed) CXOG1. (ads) + 0 = C'Dg + 201 (ads) 201 (ads) + C=O (surface) = COCL, The inactive COCl2 produced by the third step might be expected to have undergone these same reactions, leading to a chain which would have produced large quantities of inactive COs. Chemical estimations of the CO, have subsequently shown that such was not the case. There- fore if this mechanism is to be considered, a restriction must be placed on either the reactivity or the rate of formation of the phos- gene molecule formed in the third step. It can be seen that this mechanism would account for both the production of C*O, and for the exchange reaction observed. It does not account for HCl formation; however a limited amount of hydrolysis might have preceeded simul- taneously. An alternative mechanism, leading to exchange alone, would be the disseciation of C*OCl, on the charcoal surface, followed by re~ combination of chlorine with the "surface oxides" to regenerate COCl.. ~69~ This process would have caused an acewnulation of active carbon in the bed, which was found not to occur to any appreciable extent. The following additional experiments might be of value in clarifying the removal processes for phosgene: l. Determination of the amount of water actually in CWSN1LOTUS and CWSN19, dried for several hours at 150°C. 2. Further experiments similar to the ones reported, but using thoroughly dried air and charcoal dried at 150°C. or higher for pro- longed periods of time. 3. Accurate measurement of both HCl and phosgene concentrations in the effluent streams. This might be done by a combination of photometric and radioactive methods. 4. & thorough study of phosgene tube tests using standard chemical methods. The phosgene experiments, as the ones with arsine, have been limited in number; but they have yielded some unexpected results and given considerable information about the precesses studied. They have indicated that the removal process for phosgene, which had been con~ sidered well known and fairly simple, may possess unexpected compli- cations. Certainly, the results have again pointed out many additional experiments that might be performed to gain a clearer understanding of the mechanisms involved. The methed of radioactive indicators has required considerable expenditure for the instruments recuired for the studies and for nuclear bombardments. But as in its many other applications it has again proved a valuable tool for fundamental studies. Although the method is subject to limitations and must certainly be accompanied by other investigations, the techniques have proved practicable for studying gas mask absorbents. It is believed that the efforts have been justified. fat SUMMARY l. The method of radicactive indicators, using separately Ag lh , cil, and 18, has been applied in tube test studies of the removal of arsine and phosgene from air-eas mixtures by the gas mask absorbent, CYUSNLOTUS, a type A army whetlerite. Experiments were performed in which the concentration of the indicator element in the effluent stream was measured continuously, in which the accumulation of activity in the entrant layer of the charcoal was followed, and in which the distribution of activity throughout the charcoal was as- eertained, 2. A mathematical treatment for tube test absorption has been ineluded, containing the derivation of Hinshelwood's equation. 3. The deviation of the behavior in the case of arsine from that predicted by the theories was demonstrated, particularly the action of the charcoal to continue the absorption of arsine for a long time at a low constant rate. Arsine desorption was found to be negligible. 4. With phosgene it was found that nearly all the carbon in the gas penetrated the bed immediately as CO,, both with the whetlerite and its base char. 5. From the experimental Log / £2 ~ 1) funetions obtained with active chlorine, the parameters of Hinshelwood's equation were cal- culated for tube tests with varying moisture contents at ~25°C., 20°C. and 40°C, 6. In studies of the desorption of phosgene following tube tests it was found that significant quantities of phosgene were desorbed in which the carbon had been exchanged. 7. The temperature changes accompanying phosgene removal were measured. &,. Possible mechanisms to account for the results of the phos- gene absorption were discussed. (1) (2) (3) (4) (5) (6) (7) (8) (9) D. D. Ae Ae G. (10) L. Vib) J. (12) J. (13) D. (14) S. (15) L. (16) H. (17) F. Me Me Me Me Me T. M. Ne Se BIBLIOGRAPHY Yost, R. W. Dedson, and D. S. Martin, N.D.K.C. Formal Report Serial No. 175, B6; 12-16-41. Yost, BR. W. Dodson, D. S. Martin, and N. P. Nies, N.D.R.C. Formal. Report Serial No. 372, BO; 9-25-42. Yost and D. 5. Martin, N.D»R.C. Informal Report 10.1-12; b- 30-43. Prentiss, Chemicals in War, McGraw-Hill 19537 Yost, BR. W. Dedson, and Re A. Cooley; N.D.R.C. Formal Report, Serial No. 65, B6; Be 25h, Seaberg, Chem, Rev., 27, 199 (1940) . Henderson and H. W. Haggard, Noxious Gases, Chemical Catalog Go. (1927) Solomon, Phys. Rev., 60, 279 (1941) Bohart and E. Q. Adams, J.A.C.S. 42, 523 (1920) Kassel, N.D.R.C. Report, Section BG, Undesignated. Kincaid, N.D.R.C. Report, Section B6, Undesignated 9-16-43 Kineaid, N.D.R.C. Report, Section B6, Undesignated. Yost, L. M. Ridenour, and K. Shinohara, J. Chem. Phys. 3, 133, (1935). Van Voorhis, Phys. Rev., 49, 889, (1936) Shiff and R. D. Evans, Rev. of Sei. Inst. 7, 456, (1936) Davy, Phil. Trans. Roy. Soc. 102, 144, (1812) Ae Long and A. Re Olson, JdoAsGeSe 58, 221k , (1936) (18) R. S. Halford, J.A.C.S. 62, 3234, (1940) (19) J. Nielsen, Z. ges. Shiess-Sprengstoffw. 27, 134, (1932) APPENDIX I GRAPHS FOR ARSINE EXPERIMENTS. ixpt. No. Funetbion Expt. No. Function Plotted Plotted SA2—=1 D SAk=2 ,3 gh C SA2—2 L SAk~2 L SA2—3 D A SAL, L D SA2-§ L SAL@3 L SA2-6 L A SAB=1 L D D SALm~k L SAZ~2 L A D D SA3-3 L D Notes A = Accumulation of activity in entrant layer vs. time Lo it Concentration in effluent vs. time ie Lo] Distribution of activity in charcoal e t= tog (£2 -3] ve. time Thy aa a 4 of fobpebbedoged ibaa Setacs t ene vers + TYRONE IOWA Aavey eat qiet uo Swuittyie 7E-GS6 (ON CA UN (OD BSS53 7 Ta4sNeHW 797 ssf SA aon 70 JO Peery “OO HISS? V4 Pere tnt iscy “OO UV3RSI ¥ 1349NgH FreSGE ON ___ 1008 \ £4 Sma Ea — i t pane IS min. ) — —_ Z. ' { a? Aa th = i “| + 04, NO, Sol, bat Bea! ea in Ag Break af 10 cpm. ¢ doz mg /, Seni: aan renee Sane wet an es ere —t “+ +44 ae on Sein Saneinnaineneinet a es ee /8O 220 240 L200 140 /60 12a p90 FE 1022 7QAQ 62Q 5Qa0 4.020 3.00 92 6Q 120 32 20 4.0 or 4. 34. 100 10.00 9.09 g00 || 720) SOO} 1.029 t on + ' L06l: a Irw a (Br San Cwis Nie. a Cc, ap nite: ere wg, | _Ftaw. = IFA ae 7 val Temperature= 24°C. - “Break nm AY NO; So/n. em. Jo0c _ if ard i ne ee Brean a (e BED DEPTH? 2.5 Cm ee am (aok mg { ns ven Snes i ' <. i : i ; Pe ‘ Lol 4 wc eee t . 4 ‘ T ep t gq} ; He} TQ. 6.0_! $2. 42 32 al | | jyewee { et ee fe \ \ { ' | ‘ i ‘ on _ ee ee cr a os ar a nn ee 10 aaa, (Ox -/) 900 _F: L200 820 720. (2-6 = : to t 600. |}: FAQ a a 42Q vis Tue 320 a Dept th —. —_. (8-1) vs. rime . vat (ote epee 4 Carita 2 AO mg. yi . 920 € 220 122 9-0. go 70 62 SQ... 40. 3Q 4 - 4 | . an ret : ' : eas: - a —— ee _ | af ‘an . oad ed _ oot . oo — | | es re ee ; . 1 | oo i a i f £200 10.20. 120 /40 /60 180 Zz 00 100 30 ‘ “Aaéay s Ww slegeur itp: re ¥Vi-GSE ON “AUN OD B3BSU 2 WIINAH ive ~¢) 10.00 10 i-ra EESUOS WM EDRSAASEA STUNG RMD ES Gd SERN GEER CESEE EROURSONOMAREARSRESR poof Pf SAG sol i | OWS NIE TU eB foried 2) 1 ro 4QO0 fo + s00f gp Low IT gee | eb | Mattial Temperatures 22°60 | Break in AgNQ| Sola. * BOimin| | ch Break oF 10 cpm. (00238 LY9. /) | = PES ain, a A a SO SS oon sof ects een ponent pane ed — + 220 | 8 92 | 72 5a 42 22 12 : ere ees nee ae 2 1 mo an vies VR iy penen yuu apes ba fly wus rt is he ee as aa ren ae een a os Beede re Sato SHUT Want ‘ato jet ‘ACN GO wang ¥ I3ainaw (oz -/) 40.0.0 g90 f ‘goo } 700 | €20 SQQ 429 3Q0 220 J 120 go | so}. Tot 60_ i SQ ao] 3.Q_ 1 30 100 ‘yank vas JO teas TY PEXESE TON ‘ACN OS HAGGA Y 3944N3w “KaGory pau i 2ua2 a a weno ‘ sotto ~ a YieoSe "ON Va4da0aH 2 6Ge “Om "AY? “OO 43Ss% 420 /00 80 50 FO 20 pre = mag re asur une - xFAVT GOL IO ALIAILW tit wad § ¢ . haa punoiby209 BAOTO | 217 PObGOR ~ Ua i 1 (2.90 092 mestne: ean uae Kaeo & virage ‘oN WOSTAN WE MQOD ‘sus qearipesy ‘ACM COO BIGGS 4 ore iasanaye Lh POL ae 70 f O " TN ~ 7 Pegbe pon oN . bo | “8 : Spo be 7 tebe t {py rp aaee Soneeee = CWsinigs TUS - ; initial Tarp. = 2F PRC. 7 /V LOL pet. | Bréak in Ag en eae i. | i } } : \ | \ and tL +> oper a + cr ie 7 tie ; dhs { ‘ { t 10.05 pherbees 20 8.04 TAQ 2.20 90). 1.02 82. 72 62 s2_ 32 /0 a0 GO SO 40 30 "SARE Souly i penhnaeeel } FLO Se wos th co os cn oes yyountrt yy k+GSE CON ACN “GD BIGSA ¥ VIsdNIK eat: t oreBeak. v) = Bed Ca CAN. t t : Viresr ¥ APPENDIX IT GRAPHS FOR PHOSGENE EXPERIMENTS WITH ACTIVE CARBON Concentration in Effluent Vs. Time iixpt. Ho. axpt. No. {Low Flow Rate) (Noxymal Flow Rate) CG2=C CG7=LG CGh=C CGS=LC CG5=LC CG9~LC CG5=20 CGLO=LC CG6=16 CG11=1¢ CGG=-20 OGi2—G CGIA$C BOI airy 9GL O9 Os OF OF Of o/ g = — pane S mmnean Zz (SLY) = 708 I/we 96990 = MO] [- WI OLD = YAMaT peg — | (SLY) _ ci ; . / . / o98 a GO ='9 i '*}, GOFZ=ASuwaL WY _ PIltd-BPNLEIN SMI comnad | FJWIL SA LNINALT No NOGYVD FAILIV SO NOILVYALNIINODI 2D-Z9IT~ {| Jt | | | | | | Jt | | | 4 o/ Of OE lh Os o9 02 2 se WON, DAGUIDIUID “ (I LM) USEC TIN SAILN UMA Pf 00/ 06 8 a og OF Or OF oz ov bot Fo ot td T ( 1 uty) OM GO-L 1p 775 Gy 20/ = Wt 1L-0 4°f 29S yuyy 9299 = MO) 7 Lo 419 O'$ Pined Peg | fs / 7 (LY) Sou (06 1° 578i peg = (ILS) 5 Ba ="? O o 5 i i) | | PWIL SA ANFINVSAT NI De POs = CMa, “uly OWL, LE HY 0L 9 G2 WPCIBNLEINSMI NOPGYVD FAILIV SOA NOWILVYLNIINOI | | | | | I-# OD 0& os (AL Hy) el / Sap od 1} wD f9 UOMYDAMUI2UOD O02 a / oO/ FSIf{NU wy elwly WOO AOSAT penne wag’ =p7g iY _ / / _ spbwr96 2=(9-D oF bi ¢0/ = 2902 = Sway uy — OW eh OL - NY LOFG Jo FZ s1POF CNLEIN SM? ocak IWIL SA LNINIALT Nd NOGYVI FAILIV FO NOILVALNFIINOI J-/-G 9D ~ OL O92 Os OP OE Of Y/ 44 o/ Of O€ ov OF 09 02 [| jf tf | ft | Jf | tf | | | 4 DSO LOULO1LMGDIUED (PLY) HY sous), WW GILL AY ott! 7 OL o9 OF Or OL oF O/ Por or to ot bt Poy tT TT - O / — (SLY) ‘GOS *y Gd n — 7 hy IZG0 = MO/d 2908 =u, wy WIO'D = YLAId rg PIHAGD- B/SLEIN SMP FIW/IL SA LNINIASAT NI NOGAVO ~ FINMLIY AO NOILVYYLNIINOD 9-2-F OD | | | ft ty tt tt & gy S 8 z ™ N ~ S © (AL HY) VSD SJomI;/IW 2 Jo welso4puarUey /00 90 _ 40 Q K S 70 3 -, \ 60 X 9 BS XS 0 ~ 0 G — 8 30 Ls ‘2 au N £ 20 < %& & 8 % 10 0 tens oee Prats a ay | 6 6-1-C CONCENTRATION. OF _ ACTIVE CARBON IN EFFLUENT VS TIME CWS NIOTU 8 Dried . Ara Temp.= 20 “C. — C= ua TEES (5 7p) or Yors tmmelsl _. (Rm 7P) — rT TT J TT JT T T tT FF Td JOE 279 O / fr 2 Bed Depth= Soem. | / Flow = L062 "Vem sec (Rm TP) —_ es ee 10 20 30 40 SO oO 70 Time Minutes Concentration of Active Carbon - Mi 1/11 mols OY m. 3(RmTP) ee ee ee eee ee ee ee eee _ CG6-2 © CONCENTRATION OF ACTIVE _ CARBON (N EFFLUENT VS. 7/ME CWS NI9ITU8 Dried | = _ Room Temperarure=20°C. — CG ls C _ O,= 4 72—— (STP) = /0.6 mete or 102 Bed Depth = SOcm, [s0— Flow = 4062 Vem, see. (8m TP) — — — /40\— — 120 ~~ — ——~-—c, 7 /00}\— oO — 60 5 7 60|-— = — Peserpliorr — 40\— _ 20 -— =~ 0 es DG 0 10 20 30 40 50 60 702 7itmme Mi pules SOZAU IY aus Of oO! $6 OF OL o9 oS oF OF Of O/ ozs go Fra rT?Tftgkegseeretshret?set?tfdfttttT (il? = —_ N g “uonydloteg x —~ 08 3g “ bsarcncee mma g > a —or * & as _| ‘4 S * = “+ 09 RN - 7 & = SP 9 5 2 oO —00/ ~ (pzvy PF RMY y gos =meoly > — wos =Yideg Pag — N we y _o I (Luu) siram ON8 = 5 xb WF = (CLI G5 By 196 =F ozs & DeG/ = Away usy PIAD BILOINSMI IW SA LNINIGAI — WI NO@HVYD FAILIV SO NOILVALNIINOD DJ-f-2 9D | | jf | pf | | | | | Jf | | | | | ff Millimels € Ans (Ra TA Concentration of C* tor oT > fF Ff F TT TP TT L_ CGE/-C CONCENTRATION OF ACTIVE _ CARBON (N EFFLUENT VS. TIME 7 CWS NIITYUE - Dried Am. Tema = (8. 7 = ‘4 — C2 402 EES Srp) — 747 FEE or 754 mmol C 90 / (Rm 7 FP) _ Flow= /O/ “Yn yk, sae (Rm TP) Bed Depth = $0 cm — O — 80 QO 7 QO - 1@) Ne oO oO oO seman 70 — 60 — $O — 40 — JO — _ Desorption ™~ 20 |— — 10 t— coe pp tt ttttttittts 44 20 40 60 8? 100 420 142 7 ime Al inutes ) (QmTP { L961 mels C Jin s Concentration of C* 90 SO ewe yay = 30 20 f- 10 bot ror oe Eo tr tr tt i | C@9-1C CONCENTRATION OF ACTIVE — CARBON /N EFFLUENT VS. TIME CWSNIDTUB Egurl. 28°C, 7OTRH- 27 7 4,0 _~—— Oo oO O O — fe — vA Rm. Temp, =/8 °C. _ C= 952 mage (srP) = 890 mcs or 90 BROS Rm TP) | ‘rtencocead Bee Deplh= $0cm Flow = 101 2 seq (RmTP) ee Desorption 20 4? 60 80 100 120 140 Time Minutes - J 4 of C # WW tht mels Concentration (Ram TP) iS N S Q SO oe S IO 20 sammie I | CG /O--C [| CAR BON CWS N/9 Dried C= 4S /. 2 | CONCENTRATION 1 | | (N EFFLUENT VS. TIME Bed Depth= S5.0¢m _ Z (@m TP) Rew. Temp. = 18.5 °C. Cc. 79 CG a rm al © OF ACTIVE /0 20 7 IPE 30 49 ATF 79 AT OS 70 6eLPUlpW aul y G/ O9/ Ors Oz/ Qo/ oF ag Ow Of O iN rT TT T?TtqttgT — \ 3 some OF as prensa omc 8 C . x ~ 5 7 ROTH | 3 | R Q s c 9 ‘ ; = — ; “A0%/ > . LT. Y SU 990292 Ju cama > wed G9 & maO~Yv Te ome — X - at auyy dim SS = 1900000 "OG ho cperu//4waxKe By | * — 08 yy — beer _ woruriiscs 57 18 (PLOY) Ld oY EOL M7 FO 01 0.44 UO2UOD 902 MOL = Luy Wrage =4409G peg > (SL “e7/) , — — ay _ VOl{A A OF2 GF SD syousu I6l 0 Sar 9FL = | | (ALS) = 2 6¢~ =°9 PID BALE/N SMI Pp FW/L SA LNINTASI: Me NOBHVI FAILIV 410 NO/LWALNIINOD D-/-// OT i | | |} | J | | | | | | | J | | 4 Q NX a Ye CI OF a0/ ; ; (AL Ye“, yo S/O spp lid xP (PO UOVZRIAAUIIUOD Sagrulp ausy O8/ GG/ Og/ O2/ OO/ OS Og OF 02 a4 rtgstT ee ee ee ee ee ee ee ee ee ee ee UIl{fads0£3 9 — 0z — Or ~ 9 8 : A oO. — O01 O “12, ) O (JL “x) ree EO/ = Me/7 — 00/ ‘Ca? be = 44999 Peg a ° 2 G/ = 'Sual wy pee. Say SM) FWIL SA ~LNINTZST NI NOPAV2 FAILIV JO NOILVYLNIINOD 9-2/ DD J J | J | |} | J} | | | | | | | | | | (PLM) M/S spoulryp iW gD f9 WOlgos4 48902 Mil lmels / m3 (Bath TP) or c* Concentration cr | F— — F Fo Et tr t fF Ft tt LCG /4C CONCENTRATION OF ACT/VE CARBON. IN EFELUENT VS. TIME CWS NIQTUS Equi/, 25°C. 707: RH-27 Ved _. Charcoa/ Twhe & tnler Arr arn~2R§ °C. — a CG /3 C G= 203 78S (srp)= 22.2 a or 224 Ore ar (Bath 7 P) - Bead Depth = 9-0cm. —~ —_ A L_ Flow = 8.01 "7 n*, sec. (Bath TP) _ 2 80-— — > O ] % 20 x S 7 w & — £ & 160+ X _ NS i) & ° K epee by creel N S 120} — ' 8 _] ~ & ae & cs ene ead : 60 8 — * 8 = hs N _ > , 40 Q % — | | 4 Cc’ ont | [| | | | i ee ee | a] 20 y-L4] 60 82 19? 120 ‘40 APPENDIX III GRAPHS POR PHOSGENE EXPERIMENTS WITH ACTIV CHLORINE Expt. Funetion Expt. Function No. Plotted No. Plotted CG1L6—C1L ¢ CG21-1¢01 G L CG21-201 G CGL7=CL G CG21=3C1 C L CG21~1,2,3¢1 L CGLB-2E1 G CG22=C1L ¢ iL L A A CG20~1 42,361 G iL Notes A = Accumulation of activity in entrant layer vs. time ¢ = Concentration in effluent ws. time c L = log — ~ 1} vs. time OF/ sasnul ously O22 002 OF / O9/ yet et dG tr tort S S RS . S & G RS 8 RN 8 N S e) 90%. li? UIDG=YidIITD peg ; (74 bf ) —™ ea /pé jour 8 68/ “75 5a GEE =F (a5) oho J0/ =F _| ‘Do 6/ = Waal Wy PIMA OSILEOINSMI IW/L SA LNINTAZLZFI ~~ MI INIHOTHD FAILIV JO _| NOMLWVYLNIINOI /D-9/ OF | | | | [| J] | | Os S % PMLA $10 til IID wg /D JO Wolfodg #9205 SOLUS py ells f 002 I8/ OF? Oe/ 02/ ao/ oe 69 Ov oe Oo ~~ SW/L (SA a a a a a i l ™ ~ > 4+ §s ‘ ~ | or a) AN d q Is Q ‘ __ ; 9 8 ~ i °: NL > \ ~ ~. XS (AL ao) PF yyy 207) mo 7 we og = yssed Prag Ww / a. ‘7 ° /p spot bed FS / =(al ey) 53 Bul FEO = (ALS) Bum /0/ ™@ - ‘Do C1 = A4g01IV M2 PI/4IO- PSL EIN SM? 2 yo INIMOTHD FAILIV HOF (1-3 )"'907 LSI Mv7Hg 19-91 9D Jot Fo | | ft | | ft jf ft J] | |) SOLU py Obst CLF 007 oe/ oz/ 00/ Os 09 Or oz cr | | fF | | ° uy 9 ~ + 4 8 (AL Uy) 17 spo #7 bow Fu Q/ = (SLES) Ziq #1/ 29 “20S Ky MW? a9) we! 97 Fta f = aS ONY iz WIS YiIIJ peg HYL0L oF EZ /'7O8F BNL 6IN SMI r INW/L Sf LNINITSASF NI FNIYOTVHD FAI LIOv SO NOILVYLNIINGO? /92-Z/ 9D 7 | [| J | | jf ft Of OF o9 08 00/ N ~ S (PLY) 8% 77 spomrs/W we IIS GOULD HID Wag 3.0 2.0 ee ee eee ee ee ee CG ~/7 -C/ BREAK TEST — £06(2-—/) FOR ACTIVE CHLORINE VS. TIME CWS NII TW 8 EGULL, 2S °C-7ORH 220 7 47,0 — Am Temp. Z/I99T. C= 1S mac’ (STP) 7 _ | = 196 MAE? _ 2/4. AMIE Rig Bed Depth = $0 cm. Flow = 103 (ory sec. L0 _ a _ ~ f — ° \S) ely - A) 2 _ N O 4 “10 Pop t ptt tprpiy | ra] 20 4o 60 &O 100 120 7ime WKinules SPL PU) fA/ ne OPF 002 O8/ O9/ Ov! O2z/ O0/ 09 o9 Or 02 Oo a a a a a ae ae ae ee ee ee ee ee ee ee ee ee ee ° = | BS ie Ny - 8 3 0002 *. S e “ PS d > Sy & t — 0 3 s O00Pr ~\ S S PS 8 ~ & 0009 y : ode & dt feo 0 N — eS Ku fo, = Md Wragg =44dIg porg 000 0/ > -_ ad / jp faa BIZ 7e ate 9O/ = (ALS) 57 Bw S// =°9 : De b/ = PM OSIAA Df , — OTH of, OL2 HY OL R62 1'PbEF “PNL CIN SMI 000 2/ & TVOIN YH? ° SO YTAVYT AOL NO FANISOTHD FJAILIV FO NOILVIPWIIIV +: SS — /2-Z/ O99 ~ Ltt tttl ttt tt bbb SOM YY Dat Z OF/ oP/ O2/ OO/ OF O9 . CF 02 (phe) 4 Oy EO) = MOA MWIOG= 41999 Pog (SLM) - — UOnAIOSaG eid o _ [J sjouta OF) “° 55 Biz 9EL = °, — (SLI p95 bu 96% = De OF = way wy "HY, OL -HYLOL — 0 $2 NPE BAL EIN SM? ered IWIL GA LNFNISLF N/ INIYOTHD FAILW 70 NOILVALNIINOI ~19-2-3/9) fof | | tf f | | | | | | | | | “sion on O/ Og 0€ (JLweca) “/)9 spousjim y)2 Jo WolaapusrUe] re ee ee ee ee ee ee ee ee ee ee | 400'- C6 /8-C!_ = BREAK TEST — L0G ( %e-1) FOR ACTIVE CHLORINE | _ ° VS, TIME — = CWS NII-Tu 8. Equil. at 25°C. _ SO RH, 270 Te H,0 7 Temperature = 20°C 7 3.00 )— C= 7.96 Zack (Srp) ~ = = 2396 ma CG or (49 Mmmols Cl — & / 77.7 Ss Bed Depth = 5.0 crn. S — Flow = 10.3 "min sec (im TP) & — ~—d - & 200 | _ ‘ ‘ { Q re — & 7 & —~ ~~ ~~ _ > SL Je © “ 100\— q = Ss a “ — poms —_ . | A _ _ | | — 0 | | | | | | | | | | J ff 0 20 40 60 80 /00 120 (40 7/77 @ —~ NMinuates SALDPUIPY elts/ / 0OZ C8/ O9/ Ov/ O2/ OOo OF O9 Or Oz an a ae ee ee ee ee ee ee ee ee ee ee ee ee ee ee eee = — 2 = NS @ — — S ~ - 8 + § & * — 8 ~—— 000] S x S — i co Be Ss asad . S NN aa yey EO! = MO/A ~ 4 y wW2 0G =4IIg peg . | S (AL “~e) N 79 sfous was OF pDbu FEL = ~ x — / — 0002 (SLS) x55 9EL 2° < — D220 =aMwaluy _ 5S OW eh OL u ~] Ne I Oe fo He, OF JIPOT BALEINSMD boo oo ~ FINIL SA GIG AO _ YIAVT dDOLN/ FINIYOTHD FAILIW SO NO/MLVIDWIIIV /I-2-P/ OD Pet | hcl hv hE hd S09 A & S) SN S Mili rr0ls Cl /m3 (Bath TP) * NY S Corcentfration of Cl g | | tf T — 7— T — T T TT ET TT Ty yd CG ZO-12,3-Cl CONCENTRATION OF ACTIVE CHLORINE (NM EFFLUENT VS. TIME _. 0 €G 20-/ CWSNI9TUB Eguil 25 °C- 70 RH- 27 T0oho— @CGLZ0-2 CWSNISTUS Egutl 25°C SORH 7.0 TAO? _| ©CG 20-3 CWS NIDTUS Orred CHARCOAL TUBE Bath af - 25 °C. INFLUENT AIR COOLED 70 ABOUT - 20°C. C= 22./ me cs (STP) = 23.9 G&G / eri rials Cf 7, J 484 — or (Bath TP) Bed Deeth= F.O¢cm. Flow= 796 OY Ape % see. 3 _ 3 _ wn ay ss — ss _ \ Ls SS 4 \ Deserp to” ™ } [ | [| | 100 120 /40 Time Minules 4.0 3.0 ~ rs 8 2 Z 09,,(& ~ 1) For Active CAlorine S 8 ee ee ee eee ee eee _ CG -20-CI-- 2&3 BREAK TESTS — £06,(%-l) FOR ACTIVE CHLORINE ne VS. TIME CWS. N/9TUS Temperature == 25 °C _ sol Bed Depth =f0cm — lo= 22/ mteG (37-P) | = 2I9MIES or 484 AOA cf (BathTP) Flow = 796” y mx seo iN _ Q *q Q — S iS * © - Q | : _ . QS _ ; : ND | SSS & *K vwaK = . & s 2 9 S O CG-20-C/-1. Equil 25°C gs - 4 70 RH 270% HO Q — @ 66 20-C/-2. Eguil. 26°C. 7 __. SO RH. Z0 Te A,0 _ © CG 20-C/- 3, Charcoal Pried | | |} | | [| | | Jf | } ff O 20 40 60 80 100 (20 40 Time—_- Minuates Milli els, , (Bath TA of Active CAlorine Concentration fr | go Eo To For a EU — CG@2l-/-C/ CONCENTRATION OF ACTIVE — CHLORPINE (MN EFFLUENT VS TIME 200 }~ — CWS M/ITUB Egul 28°C S070RH - 2070 M0 _. Bede@rad. srtluent Keated te 40°C. —_ moe = 23.6 OBES (57 p) | = 204 mg Co or B/3 prrrols Cl ytoK (Bare TP) 7 60 Bed Depth = SOc | — Pl. — Flow = (00 sor /40 \~ (Gath TP) x N Stree ate \ peers e 120 = % — QA /00 — _ 80 _ _ O — 0 t= — 40 fe a ; 20 ) —_ ob tolbe tid” |] | | ttt st Oo /0 20 30 40 SO 60 7? 7Tipme Mrrnuates M111 melts cl 4 (Bath TP) Concentration of Cl * 186 160 Mp 120 /00 &O 60 40 20 prem: | CG 2/1-&L/ CONCENTRATION OF ACTIVE CHLORINE /(N EFFLUENT VS TIME CWSNIGTUB Equll 25 °C-7OT7.RH-~ 27 7oH,0 Charcoa/ Tube And /nkluent Hearted /n rot oF UT Bate To 40° = 235m 9G (STP)= 204 79 , “(ae th ne) Beg Deoth= 5.0 om. Flow = /00 "Uy, se. (Bath rp) | | | | Sudden Spurl of Activity r 4/3 Pot Window Vis bly Athackhed aa atoms ‘sores 20 IO 4O 7ime Wrnudres SO 60 (0 SILAW LY etllly Oo Os OL og OS Or OF Oz o/ ee ae — 7O/ Q . 78 ey DOL =MOly \ r WIOG =4YLdIq peg 2,0% £0 yg hd Peony 002 LUPIY SU) uw wegrl sOo710Y4) . (JL 9¥4°@) / cod oe 2 th om ppm E57 Bas bO2 = / . ~ (ALS) g5Bau FEZ =D ‘Deo/Z= ‘we ‘led bf PIIAD = BALEIN SMI | O0E “a. M1058 IWIL SA | | LNINISLI NI INIAOTHD FUL R | is 10 NOILVSLNIINOI /9-€6-/2Z 92 | J | | [| jf f | | | | } ff eration al $9 WOLEDIZUODUID (PL Y4eR) &' YS] 0 tet!) 1 SY J0 2.0 ~ 9 TT TT TOT TTT OT TTY CG 21-CI-~-/1 2423 LOG. » (& “if OF ACTIVE CHLORINE VS. TIME Charcoal Tube And (ntluen Bath. C= 235 ECE (STP) = L204 ng Ci er 419-4 (Bath 7 mp) | Bed Depth= 5.0 em. (Bath TP) OCG- 2/-Cl/-/ CWS NI9GTE 8. Lguil. 25°C SONDRA - MoH @ CG-2/-C/-2 CWSNI9TU 8, Equi. 25°C - 70 RH. 27070 H,0 © CG-2/-C/-3 CWS N/9Tu & Drred 1 | | | | | | | f | | | 4 Heated To 40°C sna Water mrrols Cl — Pl — Flow = 109 Cm sec. rosoneat wcrc Lr oma 10 20 IO 40 SO 60 7/fme Minusrest Milli mels/im? (Rm TP) Lloncentration of C/* 405-— FO0 }~ ¢ S | 100 r | Fo Fo Fo Fo Fo Ee PU UT CG LO-C/ CONCENTRATION OCF ACTIVE. CHLORINE IN EFFLUENT VS. J/ME CWS N/IPTYSB Orred Am. Temp.= 2/58 °C. Coe 24/ PRES (Srp) 229 aC ay 57 PROISEL Beda Degwrh = 35.0 6m Flow = @.E2Ml; Cay a SOC. Desorprion (% Break for C/* 20 40 60 80 100 120 = /40 7) me pi nures 3.0 2.0 bot oT EE EE LTT CG 22-C/ BREAK TEST — Log (£2-~);} FOR ACTIVE CHLORINE _ CVS. TIME CWS N/I TUB Orred Temperarure = 24.5 °C — Co= 223 —— CG or AT/ meets er (Rm TP) Bea Ovrpth = §.0 cm. cone < Leow = B82 mts sec. (em TP) — ec — x) Py — tS 9 — % & Q — | J} | | [ | | J J | Jf 20 40 60 80 100 120 /40 Time AN i ndres Se~ru ly -~aeWwly O9/ Ob! g2/ 00/ OP og or Oz 0 ee a ee ee ee eee 9 L— ~ — 09002 SS) = 000K Se __ Q 000 & 5 = _ 0008 cS — 5 W000! REN , —_ = > 000 2/ x3 — Sa 000 tl % P(e Le) Uol{d 40590 | 90091 BOS Me 299 =MO/7 W909 '§ = 4409G Peg 0009! “Gy baw 1% = V0 bu fez ay, wy)’9 ; — V/yzbua (bE = (Ls) 9D > — W008 Jof/Z= Mla ‘uly ; — (P8l4d) BL 6IN SMO 000 ee — GIF TWworHwVvH? —o00%e% SO SOL NI INIYOTHD FAILIV 10 12-22 DD {| | | | | [ | | Jf | | 4 | NOILVIAPNAIDVY | | | | 4 % of 4109 -b9 arog jeymusipy sad spunoD -Agiasi2y (9= APPENDIX IV TEMPERATURE MEASUREMENTS FOR PHOSGENE TESTS. Expt. No. CG7-L, S-1¢ CG?, LOC CGl1, 12¢ 0G15, 16C1 CG1L7, 18C1 Deg. C. _ Temperature 2 egrees C. Temperature ee a ee ee ee CG J-/-C TEMPERATURE OF BEO VS. TIME | CWSNIITUES Orired Co= 967209 CO¢srTe) ond O TC*#/- 46¢m. above Bottom of Bea —— a0o% CG , @ #2-2.5¢cm " ~ oe 91.0 sm nols C ¢ AmTP) © #3-0.5 . _ nT. Bed Depth = SO cm. Flow = 104 OZ, iy? AS@L, —— — (AmnTP) oT /0 20 30 40. 50 Time Niinusres. ne ee ee ee _ CG-8-/-C TEMPERATURE OF BED VS. TIME CWS NIG TUS (Oried! ~ I0 —-O = ~ 602 ECS (57e) MHeé(ght of TC JS wae, = 747m g G tion above Bottom 28 }— 25, & mimtolsC of Bed _ or as O #/ 4.5¢m 26 (Rm TP) @*2 25¢m _| © #3 O0Scm. 24 _ 22 Deserp tion, 20 4 ) | 48 Bed Depth=SOcm Flow = LOL 1 fogs 2 sec _ 6 | | | | | | | J f | | | | YY Oo 20 40 80 80 /00 120 742 Tume-~ Minygtes Temperature Deg.C Deg. Cc 7 emperature roy TOT TO UT TTT TTT TTT CG 9-C | _ TEMPERATURE OF BED VS TIME — CWSNID THES Faull, 25°C 707. RH Height of Ti 277° HMO Junction abeve = ma 66 Berrom of Bea. Cos 952 STel= 890 mace Olt: 45¢m | a 7 ©' #7 A5cem 21 Bed Depth = SOcm — Flow = 171 ml /jgise (Rr 7 P) Desorptior 20\- — oF .—O—® ~ ale o oo oN 7 ) x: | q 5 a a 4] 2o 40 60 &O /Q0 160 180 Time Minutes (Note #2 Broken) a ee ee ee ee ee ee ee ee ee ee | LCG /O- C TEMPERATURE OF BEO VS TIME — Hows NII (Dried) Height of TC — og mala, Junction Above = Com OAS /. (s TP) Bortorm of Bed. _] = 7BOL2Y or 796 MABE O*/ 45em. (Qn TP) OF3 Ofc. 30 Bed Depth =S 0c (Note #2 Broken) __ zat. Slew 10) Vemiasec, (RaTP) | 24 — © 22, - — , Manes s > ane gel 1 | 1 | Oo 10 20 SO 40 Se 60 70 Time Atlinures beg.c. Temporarure Deg. C. Temperature 30 28 26 Pr eee CG //-/-C | CWS NIG TUE (Dried) Co 84s mg CGY STP on 786mg CG; or 796 eaitnels C Vem 7P) __ ea Depth = SOcm Flow = /03 A at sec TEMPERATURE OCF BED VS TIME Pr re ee breaight of TC Junction _ Above Bottom ef Bed O F/ 45 en. @#2 25cm © #3 OF cm eens Desorption (Morst A tp) | | Jf | JO Time 60 BO MI NaS ES 1/20 [4 rT TTT TT _ C4/l2-/-C TEMPERATURE OF BEO VS. TIME CWS A199 (Oried Cox B03 KEE ( STP) = 7257/ mg or 700 mene Bey =5-0 em Flow = [0.3 1 a see (Rat?) (RmTP) Bovrom of Bed. eee | a vad erghr of TC Junctions Above on sascannnca O F/ 45 om @ #2 25cm © #3 O<5cem. {| | | | J | AA ALLE ES, Deg. x 7emperalure Degrees C. Temperature Po oT oy UT | i CG] IS5-1-C/ TEMPERATURE OF BED VS. T/ME CWS NI9DTUS (Oried) Flows /O3 7 at a A Te Break C/* ree * Sc | e *2 45¢m * © £3 A$ cm *" mglG or | Bed Débrh = $0 cm. C,= 852 mach srp)= 7 a/ o7C W - 45cm abeve boftom of oe [S58 rere als Cl wn Bed cement Oo 20 40 60 120 [4-0 Time Minutes ee ee eee | r to ot hg CG (6-/-C/. - TEMPERATURE OF | B8E8| VS. TIME CWS NII TUS (Pried) —_ _ mg Gb. Helgh of FC Cy = /0/ (OTP) molscl Junction above = 9.38 T2S5°% oy 189% a — (Rm TP) Bottom of Bed. —e j fy 3 _— Bed De th = £0 ° oe Seem pth = $0 om. @| 72 25cm FO |— Flow (02 Yt, see. © #430 O05 tm 7 28 __ x { 261— /%o Break a —-© O-» — for c/* | e* i | N, ~ pe Desorption L | t | | 60 80 430 ‘FO ‘70 Tima Ad »' un Jac Deg. C. Temperature Deg. Cc. Tem perarure rT oT FT FT T TT FF Yo yo ot Et UE UT Yd _ CG /7-C/ _ TEMPERATURE @F CHARCOAL BED VS. TIME = CWS NW/3S TUS Egur/. 25 "0 70 RA. 2ROVEHZO _I ‘ O TC #/ Suncetien &£em Above 240\— Bed Deprh = ye CHT, Bottom of Flew= /03 my 3 Sec Bed (Ren TP) @e7c *2 Z 25 emu _| 200 ry 1%. Break» . ee x a D | Desorprion. 180 a —O—O | —_— 169 | Go™ ys rps sar jas cg (Rm T-P)| " i molt 1d ir ioe |) | Ae 0 20 4 100 —" (90 40 Time Mrnutes re ee ee ee ee ee ee ee ee ee ee | = CE (8-Cl-/ _ TEMPERATURE OF CHARCOAL BED VS TIME — EWS NID TUB Eguil 25. SOTRH, 7070 HO _. 22 be Co= 796 m ms C6 (STP)= 736 ang CS 4, /49 mem el$ EL py 7 9° ae Bed Depth = S Oem. Rm Temp. = 21°C. 7 Flow = 10.3 "Vem? « sec. (Rm 7P) ray _ 19 0.76") Junction 45cm Abore- x Bottom of Bed. /8 e7e 2 * 25" — 7 © 7c*3 OS 16 _ IS — a~@ctltttiittpttiy oO 10 20 30 40 £0 60 Time Minutes Propositions Le The Navy should modify its present high altitude oxygen rebreather to incorporate an automatic nitrogen purging pump and pressure relief assembly in the breathing bag. 26 The kinetias for the elimination of elementary nitrogen in the human body may be defined by the differential equations dx ’ i= ke, (@yp=x) SY = ky (aop-y) ae 2\8gpr¥ where x + y = Amount of nitrogen dissolved in the bedy fluids and tissues. 6 = Time, p = Partial pressure of nitrogen in inspired air. Tentative values of the constants may be evaluated from Behnke's™ experimental work. The solution of these equations, when combined with the proper condition equations, permits the calculation of nitrogen partial pressures in rebreather apparatus under a variety of conditions. * Behnke, U.S. Nav. Ms Bull. 35, 219 (1937) 5e A more useful expression for the response of a counting rate- meter than that given by Spiff and Bvans™ is f(t) = KCL, & RO a) 6 where f(t) = The counting rate Ip = The current through the tank curcuit resistor % a» Time RG « The time constant for the tank circuit * Spifrf and Bwans, Rev. of Soi, Inst., 7, 456 (1936) deo The "temperature correction” factors, provided with the drmy Air Portes oxygen test kit of the Scholander fype™, are for sea level pressures only, A simple addition to the correction formula would per~ mit the apparatus to be used at low pressures and would sroid confusion. * NRC Report, CAM 757, P. F. Scholander 5. The analysis of arsinc-air mixtures with the “conventional” iodometric methods is unreliable because of the slow ebsorption of arsine by KIg solutions. However the portable apparatus of Tutwiler*, developed for the rapid determination of HS in coal gas can be modi-~ fied for analyses of mixtures of interest to charcoal tube testing, 5-10 mg/L, to give an accuracy of 12%, * gutwiler, J.AC.5., 25, 178, (1901) 6s For the assumed kinetics equation for charcoal tube testing™, the equation is proposed o ii ot* a(c=bW) where Nw The conentration of material on the absorbent @ = The concentration of the toxie gag in the air. % = Time a,b Constants A simple solution of this equation from which the parameters could be easily obtained might prove of value for correlating Gharacoal per- formance. / * thesis 7. The use of an “ideal” closed circuit breathing system for study- ing Gases of radium poisoning would permit the measurement of higher radon concentrations, the order of 20 times greater, in the expired aire 8. The techniques reported for the measurement of vapor pressure™ of D0 have not possessed an accuracy consistent with the purity with which the material may be prepared, If accurate values are needed, further determinations mist be made, possibly using a direst meagureuent rather then a differential tensimeter. * Lewis and McDonald, J.A.¢.S., 55, 3057 (1933) Wiles and Mengies, J.A.C.8., 68, 1067 (1936) Riesenfild and Chang, 2. Phys. Chem. B33 120, (1936) 9. The static method of Iwage and Sano™ for measuring the equili+ bria of HyQ and Hy over @ metal-metal oxide system does not eliminate thermal diffusion, contrary to olains. * Jwase and Sano, Sci. Rpts. Tohoku Imp. Univ. 1936 465 10. Teaching assignments for graduate student assistants in the Division of Ghemistry should be arranged to give, se far as is pos~ sible, each man a year with an organic chemistry course, a year with a physical or analytical chemistry course and a year with at least one recitation group.