Патент USA US2404887код для вставки
«MV 3» §94- R. E. REEVES Erm., ` ' l ' , STABILIZATION OF NI'.I‘ROCI'BI'JLULOSE` 2,404» 87 \ Filed May 5, 1944 ` Ely.; _El 4.0 -î 3.0 a_PsmMocEmLgu4osël 0.0 0 l0 2O 3o 4o 5o ` STAEILITY (immun-:5) l 3&3 . > l - ì . ad“ ' ' amm/má 2,404,887 Patented July 30, 1946 UNITED STATES PATENT OFFICE STABILIZATION 0F NITROCELLULOSE Richard E. Reeves and Joel E. Giddens, New O_r- _l leans, La., assignors to United States of Amer ica as represented by Claude R. Wickard, Sec retary of Agriculture, and his successors in» of-` Application May 5, 1944, serial No. 534,370. 2 claims. (01260-223) (Granted under the act of March 3, 1883, as l. amended April 30, 1928; 370 0. G. 757) V l ~ This application is made under the act of March 3, 1883, as amended by the act of April 30, 1928, and the invention herein described, if patented, may be manufactured and used by or `for the Government of the United States of America for governmental purposes without the payment to us of any royalty thereon. This invention relates to nitrated cellulose, and has for its primary objects an improvement in the process “of stabilizing such cellulose, and such other objects as will be apparent in considera tion of the following specification: Due to the use of sulfuric acid in the prepara tion of nitrocellulose, it is ordinarily necessary to remove the retained sulfate to stabilize the product. This is conventionally accomplished by repeated boiling in water and beating treatments, a long and tedious process. It is also known that addition of a little ammonia to the boiling water aids in stabilization. However, in this event it is necessary to control the amount. of ammonia added very closely, as the addition of , ammonia little beyond slight alkalinity deleteriously affects 2 o ’ linters, and is also effective with nitrocellulose derived fromwood pulp. l . _ _ _ _. _. The process is more specifically exhibited "nf, following examples:EXAMPLEI .;_ A ~ ~_ , > __ Cotton fibers, cut to about 0.15 inch and> purif‘, ñed to meet the specifications for linters to be', used in the productionf-of‘fpyro’.’ nitrocellulose' by the Naval Powder Factory, were dried over@v night at 105° C. A `32 g.- portion oftheldried» iiber was stirred in a solutionconsisting of,v 231i> g., of H2O, '866 g. 0fA H2SO4, andÍ 494 g. QÉHNOS, all“ chemicals used beingvof C_rlP. or reagent grade. The mixture wask maintained at _40° C. for _2Q minutes,v immediately after which the formed nitrocellulose was ñltered and drowned in cold tap water, It was then washed with 1l) changes of water, and was boiled f_or 4 hours underreñux _and dried, giving a lot having a nitrogen cont The boiled nitrocelluloseas. above, obtained. was divided in four parts, one vof which was rinsed and dried and kept as a controlsample. The the nitrocellulose. According to this invention, ammonia isA used at 25 other threek parts were separately stirred. for.¿.1-_5_ minutes, each with 50 parts o_f ammonium hy@ a lower temperature, preferably room tempera droxide solution containing 0.05%, »0.20% and ture, as one treatment, in which event the alka~ 1.00% of ammonia, respectively, at roomtempera-Ä linity need not be so carefully controlled.` Also.4 tures„to obtain three separate stabilized samples'.v ammonium salts may be used instead of the hydroxide, andif the treating solution is buffered to remain slightly acidic, excess salt can be used with a hot solution with no deleterious eiîect on the nitrocellulose, and thus be introduced in con~ nection with a boiling treatment as claimed in copending application, Serial No. 670,086, filed May 16, 1946. . Preferably, the ammonia treatment is combined with at least a‘prior boil, and boiling and beat ing treatments may be advantageously employed subsequent to the boil but prior to the ammonia treatment. Room temperatures are most satisfactory for the ammonia treatment because of convenience, althoughhigher and lower temperatures are effec tive. _In general; it may be said, however, that the higher the temperature the more closely the alkalinity must be controlled. The` process has been carried out principally using cut cotton as the cellulosic raw material, but The excess ammonium hydroxide was removed Tests were run on these samples against-the control sample using the known134.5° heat'test (Military explosives, War Department TechnicaA Manual 9~2900. August 29, 1940, DD. 66 andö'l), and also using a `second procedure'hereinafter designated as the ‘_‘110° test.” This second _pre-_' cedure was developed especially to handle small samples with safety and reproducibility, and is illustrated by the apparatusïshown in Figure 1 of the drawing. In thisapparatus, -.a.„suitable quantity (0.5 g. with _an apparatus having dimens sions as shown in the drawing) ofthe sample to be tested is placed in a U-tubeläl immersed -inïa tricresyl phosphate bath fl! maintained at 110°i0.55’ C. 'The U-tube IB isstoppered at its` upper ends, and has one of its legs connected by a tube l2 to one leg of a second U-tube I3. An indicator consisting of l ml. of an aqueous solu-ï has been found’to be equally effective with cotton 50 tion of 0.5% potassium iodate,__1.0% potassium 2,404,887 3 4 iodide, and 3 drops of a 2% starch solution is ing treatment, the other part being stirred for placed in the second U-tube I3. This U-tube 15 minutes at room temperature in 1%V aqueous is also stoppered at its upper ends, and its other ammonia following the boiling and beating treat leg is connected through a tube I4 to a suitable ments, rinsed and dried. suction device (not shown). Outside air is drawn CY In the boiling and beating treatment, the water from a tube I5 at the rate of 30 ml. per minute was changed at the end of 4, 16, 28, 40, 50, 63 and through sulfuric acid in flask I6, thence through 70 hours. soda line in bulb I'I, thence passing through Test results on the removed samples are given U-tube IIL sweeping through the sample therein, in the following table, the asterisks in the table and thence passing through tube I3, bubbling indicating that the samples were too unstable to through the indicator therein. The air sweeps be tested by the procedure employed, and the the volatile decomposition products from the :blanks indicating that no test was conducted in sample, the end of the test being determined when` the particular instances. blue coloration develops in the` indicator. The Table Z time required for this coloration to'occur is taken as a measure of the stability of the sample tested_ With the 110° test, as above described, the sta 134.5° heat test 110° test bility of highly unstable nitrocellulose samples may be evaluated with comparative safety and the reproducibility with such samples is excellent. 20 Boiling and beating treatment . . Control Atlëèä‘êxàla Control Aälelzïllt‘èrâm Moisture control of the sample is not such a critical factor as in the case of those tests which part’ part Part’ part mmutes minutes mmutes minutes keep evolved gases in contact- with the heated sample, and therefore the 110° test may be used None., _______________ __*_____ with air-dried samples. 25 The 110° test and the 134.5° heat test often Boiled 4 hrs. Boiled 14 hrs vyield approximately the same times for a given sample. In both tests samples showing times of Boiled 2 hrs. * 20,20 __________ __ 25,25 26, 26 Boiled 2O hrs _. good stability. 18,25 29, 18 ' 35,30 25, 25 __________ __ 5, 5 27, 31 Boiled 30 hrs ........................... „_-_.. Boiled 40 hrs ....... __ 15, 17 32, 37 8, 7 7, 7 32, 26 23, 17 27, 23 15, 23, 20 Boiled 40 hrs. and beaten ___________ __ 26, 24 less than 10 minutes may be classified as having Boiled 50 hrs. (beaten poor stability, those showing from 10 minutes to 30 at 40 hrs.) __________________ __ Boiled 63 hrs. and 25 minutes as having moderate stability, and `those showing more than 25 minutes as having 2,2 4, 4 4, 5 4, 5 beaten (2d time)___ Boiled 70 hrs. (beaten at 40 and 63 hrs.)_-_ 23, 23 23 ________ __ 16, 22 27, 29 40, 40 20,23 32, 30 24, 24 24, 24 16, 14 23, 23 The following table indicates results of the tests of the control and the three samples of Example I, 35 Comparison cf results given in the data of Ta the several results indicated from each test of a ble 2 shows that ammonia treatment produces a sample being obtained by repeated tests on the greater change in stability inthe earlier stage of same sample. ‘ Table 1 the boiling and beating treatments than in the Y later stages. A product of high stability is ob 40 tained if ammonia is used with only 4 hours prior Treatment 134:.5‘x heat 110° test test, minutes y minutes boil, whereas without the use of ammonia a 40 hour .boil is required to give a corresponding de gree of stability. 7 3, 4 0.05% NH3 0.20% NH Control. _ __- 32, 29 30, 30 ' 80, 52. 64 40, 69,35 1.00% NHL.“ 37, 35 102,100, 72, 60 In order to- determine whether the stabilizing effect of' the ammonia treatment isdue to the y alkality of the ammonia, experiments were con ducted on similar samples of freshly prepared nitrocellulose using dilute aqueous solutions of other basic` substances such as lithium carbonate, ’sodium ' hydroxide, methylamine, etholamine, pyridine, and urea. The stabilizing eifects of these were nil. Apparently, therefore, the stabi lizing action of ammonia is not due to its alka linity, or at least due to this alone, but involves other properties. As illustrated by the following example, im provement in stability is obtained by combining the ammonia treatment with a variety of boiling and beating treatments. EXAMPLE II A 64 g. batch of cut cotton was nitrated in a manner similar to that employed in Example I, giving a lot having a nitrogen content of 12.66%. All boiling and beating treatments were made in distilled water and with a Waring blender. At various `stages of the boiling, with or without beating, Samples were removed and each removed sample was divided into two‘parts. One part was rinsed and dried and used as the control to indi cate stability derived from the boiling and beat. ' ‘ In the examples given above, the ammonia treatment was eifected by using dilute aqueous ammonium hydroxide as the stabilizing agent. However, other ammonium agents, such as am monium sulfate, may be used. For- example, freshly prepared nitrocellulose may be boiled with a slightly acid solution containing ammonium salts buffered to remain at a pI-I higher than 3.5, preferably higher than 4, with excellent stabiliz ing results. In this case, the concentration of ammonium salts in the solution may be as low as 0.10%, calculated as ammonium sulfate. The buffering agent may be any which will give a satisfactory buffering action in the faintly acid region. Sodium acetate', potassium acid phthal 60 ate, and sodium citrate have been found satisfac tory. One hour lboiling in such a solution gives a nitrocellulose of excellent stability. Cut cotton fibers were used in the preceding examples. The treatment with cotton linters 65 gives similar stabilization results, as illustrated in the following example: EXAMPLE III A batch of nitrocellulose prepared from cot ton linters and having a nitrogen content of 12.73% was divided into two parts. One part was boiled for 4 hours without ammonia treat mentV and the other was placed in water, adjusted to slight alkalinity with ammonium hydroxide, and stirred for 1 hour at room temperature. 2,404,887 , 6 5 nitrocellulose, essentially the same procedure of Test results on the two parts are shown in Ta extraction may be followed as in analysis for the bleu 3. sulfate. In this case, it is better to use the 0.01` normal I-ICl solution for the boiling. The re moved ammonia may then be determined by use of Nessler’s reagent (see formula of Bock and Table 3 Treatment ls‘î‘tîsfeat 110° test Benedict, as given by Hawk and Bergeim, Prac stiriedihburin dilute NH, ......... 2o so, so tical Physiological Chemistry (1937),` 111th ed., Blakiston, Philadelphia, Pennsylvania, p. 928) and 10 the blank and test samples compared with the .Coleman spectrophotometer at 425 mu, the quan-V As before stated, the action of the ammonia is Boiled 4 hours, no NH; ______________________________ . _ ` ` 3, 3 not due alone to its alkalinity, as has been amply tity of ammonia being read from a standard curve demonstrated by substitution of other bases. vAlthough the true action is not definitely known, established with known solutions of ammonium chloride. i . « .. g _ By using the 0.01 normal HC1 solution for boil it may be stated that it is probably due to an 15 ing, both the sulfate and ammonia determina-1 ability of the ammonia to penetrate the structure tions may be made simultaneously on the same of the nitrocellulose fiber and neutralize thesul sample. ` \ ` V f fate bound within the ñfber structure or within In analysis for the ammonia according to the the molecules themselves. Experiments relative to sulfate contents and consumed ammonia con 20 above-outlined procedure, ithas been found thaty repeated boiling with changes of solution does tent tend to exhibit this. Such experiments also not show the amount of ammonia removed in indicate the consumption of ammonia in the successive extracts to approach zero, as would process and optimum conditions relative to other be expected. Rather, the readings approach a factors, and they give an indication of the nature of the sulfate association with the nitrocellulose v25 minimum of approximately 0.03 mg. NH3- per g. of nitrocellulose. This minimum is usually structure. reached after four boiling treatments. , ' Quantitative analyses of sulfate and ammonia Testing the analysis on 70 treatments with‘boil in nitrocellulose present numerous difficulties. ing and beating, using many samples of nitrocellu The amounts of sulfate to be determined lie be tween 0.1 mg. and 8.0 mg. S04 per g. of nitro 30 lose which had not been exposed to an ammonia treatment and were presumably free from ammo cellulose, .and the ammonia contents vary be nia, showed an average of 0.034 mg. NH3 (or at tween 0.03 mg. and 2.0 mg. NI-Ls per g. The anal least a substance which reacted with Nessler’s re yses which involve the destruction of the nitro agent), per g. of nitrocellulose.` Therefore, it has cellulose by chemical digestion are not trust worthy under these conditions, as they introduce 35 been concluded that the observed minimum of ap proximately 0.03 mg. NH3 per g. of nitrocellu complicating factors obscuring the desired results lose is a systematic error in the procedure and when dealing with such small percentages of sul all values are adjusted by subtracting 0.034 from fate and ammonia. the observed value in each separate reading. A substitute procedure involving mechanical It has also been found that the vigorous beat-_ separation of the sulfate and ammonia was used, 40 therefore, in which comparatively large samples ing given the nitrocellulose in the above-described manner of mechanical separation of the sulfate, are given repeated boiling in pure water or in a quickly liberates the sulfate even from samples dilute aqueous solution of HCl and beating treat which had been subjected to long boiling, and ments. In this manner, extracts are obtainedY which contain a minimum of interfering decom- ~` that a short period of >locating (5 minutes) is accompanied by an increase in stability of the position products from the cellulose nitrate. By ' using suitably large samples and by concentrat nitrocellulose, provided the nitrocellulose has al ready been extensively boiled. However,> con ing the extracts, it is possible to analyze for the tinued beating introduces unstability. Therefore, very small traces of sulfate and ammonia re it appears that the beating results in two com moved from the nitrocellulose which probably peting factors: first, stabilizing due to separa escape detection in the usual methods of analysis. tion of the sulfate from the nitrocellulose, accom The vigorous beating action desired in this plished in a short period; and second, a harmful method of analysis is well obtained by use of a ` effect probably due to some action of the beater Waring blender. on the nitrocellulose itself, introducing unsta For analysis of the sulfate, 58 hours. boiling bility in the longer periods of beating. For this with eight changes of approximately 100 parts reason, unless otherwise noted, all described beat of distilled water or of a 0.01 normal HC1 so ing periods are for 5 minutes. ' lution of the sample per part, may be employed, By using the extraction method of analysis for followed by one beating. The severa1 filtered ex tracts may then be combined for sulfate deter (il) the sulfate and by testing the nitrocellulose for stability at various stages in the extracting proc ‘ mination or each may be investigated separately ess, precise information on the effect of sulfate and the amounts totaled for determination of the on stability was obtained. As' representativaa total sulfate content. Determination of the ex 20 g. sample of nitrocellulose (13.0% N) 4was tracted sulfate is readily accomplished by the boiled with changes of distilled water, beating turbidimetric method of` Treon and Crutchfield treatments being introduced at intervals, and the (Ind. Eng. Chem., anal. (1942) ed. 14, p. 119) several extracts were analyzed for S04, the proc modiñed to allow for use of smaller volumes of essbeing continued until a negligible amount ofA solutions. With a 25 m1,'volume, ‘amounts of S04 SOiwas'removed in the >final' treatments.- By between 011mg. and 2.0 mg. may be determined computing the amount of 'sulfate removed in each with considerable accuracy. Observations may be made in a rColeman Model 11 spectrophotom eter at a wave length of 540 mu and the values read from a standard curve established with known solutions of sulfuric acid. ‘ In determining the ammonia content of the extract per gram of sample and adding the re-- sults, the total sulfate content of the original sample was obtained. Also, the sulfate content of the sample at the beginning of any stage of removal was similarly determined by adding'the, _ 2,404,887 8 S04 .determinations of theextracts including that lulose (12.82% N), treated only by rinsing, which stage and those following. >Table 4 indicates the was then stirred for 15 minutes in a 1% aqueous ammonia solution and washed with distilled water until the washings gave no test with Nessler’s re‘-` results of lone such experiment. ^ ,Tabfzç 4 5 agent. This sample had a stability of 45 minutes , . f T H ~ ' 1f t'i _ , with the 110° test, as compared to2 minutes in 'rated o aoîiliiî e stability’ the absence of the ammonia treatment.. ,.Ten.v boil, Ssmpïe’at 11Q°iest grams of the ammonia-treated sample _was re, h‘m’s ‘îßeigf'lïgîg muws i f Treatment . Rinsednu-__________ __'_ ____ ,V goileâï? ärs- ~ Bghâd Íôáîg 0 - g 23 .L05 __________ __ ñltering on washed paper were made up to known 0184 '4 volume for determination of .sulfate andammo-V nia contents. Toward the end of the boiling, .a g beating treatment was introduced. The amountsA ggägä îâlrîíhlgîêz eaten 10mm ““““““““““ " 0125 Boiled ,7,1m ---------------- ~~ Bäàlêîläâlëâ‘ìf, ägägg Èôhïrlës- 53 ' 0-20 gg gjîg 8~ _____ -~ Boiled 1c 135111111111 97 12 l5 of sulfate'and ammonia removed fromthe sam ‘ 15 ple by the various extracts are indicated in Table ‘ Ígjgg 16 5, all ammonia Values being adjusted by sub traction of the 0,34 mg./g. to compensate for the -olia """ "ëöïèò Boiled 58 hrs. with 8 changes f ci water and beaten 5mm.“k 'peaiediy boiled with 25o m1. `ef aornermai HC1, l0 and the extracts obtained from each boiling by 155 systematic error before described. 20 0.00 __________ _. Table 6 The relationship of sulfate content and stability 1s morel easlly vlsualized by reference to 4 , ‘ . _ - - Y Boiling treatment . . Ammonia' Sulfateœo) moved’ removed, l ing/g. mgJg' . re- curve A shown in Figure 2 of the drawing, in 25 ` 4 which the ordinate represents S04 content in mg. per g. of nitrocellulose, and the abscissa rep- 1st extract (boiled 14 hrs.) ........... __ 1-08 4.35 resents stability as determined by the 110° test iââìiíäëìÉt‘âiiìäÈtïìììiiiiiïïïï: It? 13% the curve indicating less stability after the 30- 4th extract (boiled2hrs.) ............ __ V.00 .15 fggi‘llâgàäfgìf; -01 .18 minute is introduced point, due sincetounstability the beatingbeyond treatment. that point 30 Curve A dirigere 2 indicates that stability is '0°' a function' of S04 content, provided other fac- Total ------------------------- -. tors, such as excessive beating, are not introduced . ‘m ,1-40 V k6.43 _ which tend to alter the stability. 35 The data of Table 6 show that as` regards the If ammonia treatment is given, the Stability sample tested, 1.4 mg. of ammonia is suñicient to results are entirely different, as indicated in Tastabilize l g. of freshly prepared nitrocellulose ble 5 showing tests on a number of samples of containing 6.43 mg. of sulfate. This gives aratio nitrocellulose having dilïerent S04 contents, and of mols of NH3 to Inols of S04 oi' 1.20. treated for l5 minutes at room temperature with 40 A number of other‘experiments similar to thel a'1% ammonia solution. one described next above were carried out with Table 5 other samples, some without and some with the boiling treatment, the results of which are shown SO C011tent 1%treatment ammqnia . V Pmwustreatmentufsamples _ es ’mm es mused free ofaeids .4.64 Boiled 30 min-.. 3.20 , 28 Boiled 2 hrs_,_ 1.40 25, 25 Boiled/8ms.. ggâlgâ ller: _ _ _ _ _ _ V.a gî'labble 7, e , an firtsä lirlieinesbeiëigeingthecorrespon total linging of e o i er . Digg/21,65’ ìtaëüity 1ì0° à” lines for the other samples, a column for the _ s .67 _ _ _ _ _ _ _ _ __ eeeeeeaehaitiana-attimi: 1415,18 Table 7 35,45 5() fg, :di ratio of meis ei NH3 te mois di S04 being added. 15118 The results of Table 5 are visually represented Treatment pil'leceding stabiliza- “m W“ 1% mmm Rinsed (from Table 6). in curve B of Figure 2. This curve shows that 55 äfilïse‘äìáógñfjï‘m* good stability is achieved by the use of the am- Total Total amrämia sulfate (ILL/3g); _ 1.40 :_ B0i1ed2hrs ___________ __ _____ __ .37 Mol NH.; tig/4g); . M015@ 6.43 »1.20 Éjgâ ljââ 1.40 L49 momia treatment with nitrocellulose containing as » high as 3.2 mg. S04 per g. of nitrocellulose. As the S04 content approaches zero, curves A and The data of Table ’l show toc wide a variation to rbe taken as establishing an exact vrelationship B tend to meet, indicating that the effect of the 00 between the ammonia consumed in the treatment ammonia treatment ls to oiïset the unstability due and the sulfate present in the sample.` However, to presence~ of the S04. Also, apparently the arnin the tests for determination of ammoniaQthe monia treatment has no elîect on other factors ammonia content in the sample will include not causing'unstability. only that consumed in neutralizing the unstabili , Again, it is mentioned that the stability ob- U5 zation dueto the S04, but probably in addition -a tained by the ammonia treatment is not due, sole. ly atleast, to its alkalinity, since tests with other _ portions of the same samples used in obtaining the data for Table 5, but substituting a sodium hydroxide treatment for the same small amount bound' by the nitrocellulose within the über structure, held by liquid retained on the treated über, o_r both. It is‘probable, therefore, that the ratios are higher than the ratios required ammonia 70 for neutralization of the S04 itself. treatment, gave no noticeable increasein stability over the samples not so treated; The marmer `by which ammonia brings about stabilization of nitrocellulose is indicated in the following experiment using a sample of nitrocel- 25 - « . If the sulfate is present in the nitrocellulose über as half-ester represented by R-S04H as has been suggested, l mol of NH3 would combine With l mol of S04. If, however, the sulfate is.l present' as free acid, as has been suggested'also, 2 molspf 9 2,404,887 NH3 Would combine with 1 mol of S04. There process is attended with little work and. is accom fore, the data above favor the half-ester sugges plished in a short time, it may be seen that the tion. invention involves a very simple and inexpensive In another experiment, freshly prepared nitro manner of treatment. cellulose (12.6% N) was treated with ammonia, 5 The above examples and experiments were all rinsed and dried, giving a resulting stability with conducted with laboratory-prepared samples and the 110° test of 18 minutes. Three grams of the equipment. In the following example, the nitro stabilized material was then stirred for 10 min cellulose sample was from a batch being prepared utes with 75 ml. of 5 normal HCl, rinsed free from for `making pyro powder. Cl ions, and dried, resulting in a nitrocellulose re 10 EXAMPLE IV duced in stability to 5 minutes, which could not be improved by further rinsing. However, upon 1.5 pounds of nitrocellulose were taken from a a retreatment with ammonia, good stability again boiling tub following two 4-hour periods of boil obtained. ing with changes of water. The sample was The evidence at hand as shown by the fore 15 beaten in a laboratory beater to ñneness approxi going experiments indicates that the ammonia mating that of ñnished pyro. It was next boiled treatment oiîsets the harmful eifect of the sul« for four hours and allowed to settle. The water furic acid bound within the structure or fiber of was decanted, replaced with cold water to which the nitrocellulose, and that other bases, at least was added aqueous ammonia containing NH3 the many tried, are unable to accomplish this re 20 amounting to 1 percent of the weight of the ni sult, possibly because of inability to penetrate to trocellulose. The suspension was stirred at room the position where the sulfate is iixed. temperature for 1 hour, allowed- to settle, and It may be noted from the experiments that a washed by decantation with eight changes of very small amount of ammonia actually is in water. The product gave stability of 29 minutes volved in the stabilization of the nitrocellulose, 25 by the 134.5° German test and 37 minutes by the 1.40 mg. of NH3 per g. of nitrocellulose being in 65.5° starch-potassium iodide test (this test is dicated in Table 6. It has been shown that with a standard procedure described in War Depart nitrocellulose which has been boiled for 4 hours ment Technical Manual 'I‘M 9-29900). Similar and beaten, less than 0.1 mg. per g. is required. stable products were produced with 6.5 to 15 These amounts are almost negligible. 30 hours of boiling. In actual production of the stabilized nitro It is desirable that a certain amount of boil cellulose, the ammonia consumption would in ing follow the beating treatment. The beater inv clude also the small amount bound by the nitro this example was a. small paper mill type beater. cellulose and any lost by the liquid retained on Excellent stability was also produced when am the ñber when draining the treating solution 35 monia was introduced into the solution during from the nitrocellulose. The liquid loss involved in this latter amount, without taking special precautions for recovery, may be roughly ten times the weight of the nitrocellulose ñber the beating operation. This has the advantage of reducing acidity and hence reducing the cor rosion of the beater as Well as produing good stability. treated. Thiswould involve a loss of 5 pounds 40 Having thus described the invention, what is of ammonia per 1,000 pounds of nitrocellulose claimed is: on the basis of employment of a 0.05% ammonia 1. A process comprising stirring nitrocellulose solution. Under these conditions, the total loss in an aqueous solution containing ammonia- at of ammonia would approximate 0.64% of the room temperature for a time not to exceed about weight of the fiber treated, and this could be con 45 l hour, thereby to stabilize the nitrocellulose. siderably decreased by pressing or centrifuging 2. A process comprising stirring nitrocellulose the treated fiber to recover the retained liquid with weak ammonium hydroxide at room tem or by recovery of the ammonia from the water perature for a time not to exceed aboutl 1 hour, used in washing the retained liquid from the ñber. thereby to stabilize the nitrocellulose. Since the ammonia solution drained ou' the 50 batch of treated nitrocellulose may be reused, if desired, in treating other batches, and since the RICHARD E. REEVES. JOEL E. GIDDENS.