THE pH AND THE CONDIUOMS 230R ERECHEETATECW 055 SOME MWALMC HYS‘ROXWES EN CONCENTRAWD EUFFERED NICKEL SULFATE SOLUTIQNS Thasis for film fiegme as! M. S. MiCHQGAN STATE COLLEGE Wafia Charms-3: McCiuggage fiQé? n ' Q - Thisiltoeertifgthatthe thesis entitled 1L. g... «ML :1. came... 7:» M 4m M‘JMM Edam.“ M CWAJT-t) WC) .>(. ,4. o—Quab 5-“ , presented by W04». MAW W°Wn°v has been accepted towards fulfillment of the requirements for degree in (WJ) 237m fiajor prof DateW- The writer wishes to express his sincere apprecia- ticn to Dr. D. T. Ewing for his guidance and suggestions which made this work possible. THE pH AND 1‘ 1; CCIIDITI’JNS FOP P7"CIPI"1I‘.M‘J_' N OF SCLE itiL'I'nLLIC HYDIC . ES N CJN’ILvTi.tTflD UFI‘T‘FII NICLEL SULFATE SCI? IONS LY VxADE CHARLES XCCLUGGAGE ’ —-¢ A‘TdLSIS Submitted to the Graduate SGhool of Michig State College of Agriculture and Appliedan Science in partial fulfilment of the requirements for the degree of MASTER OF SCIENCE 1949 WY DEPT. "“"A'f' '77 .L w fl) ‘ (4" M (A? ‘7 HE pH n33 THE CSNDITIONS FOR PRECIPllaT SN 0? SOME NETaLLIC HYB?OKIDES I? CONCENTRATED BUFFERED NICKEL SULFATE SCLUTICNS A study has been made of the factors effecting the formation of the hydroxides of aluminum, cadmium, tri- valent and hexavalent chromium, cupric copper, ferrous and ferric iron, and zinc in a strong solution of buffered nickel sulfate with the object of determining the pH values at which the hydroxides began to precipitate, and at which they were completely renoved from solution. Various pH studies of precipitation of metallic hy- droxides have buen reported in the literature but the only previous extensive investigations similar to the present one are those of Hildebrand (l) and of Britten (2). atkins (5) found colorimetrically that ferrous hydrox- ide precipitated from ferrous sulfate solutions between pH 5.1 and 7.6, while Patton and Mains (4) found the range to be mcre specifically between pH 5.5 and 6.0. Pickering (5) ObSGPV:d that complete precipitation of ferrous iron had occurred at the pH where the solution was alkaline to phenophthalein. Miller (6) investigated the composition of the precip- itates formed from alum solutions upon addition of alkali. A pH of 5.5 (colorimetric) was given as the value at which the precipitation of aluminum approaches completion. Blum (7) 21761.8 (0/ ,/ studied the precipitation of aluminum as oxide and from observations made with an hydrogen electrode and suitable indicators, for example, methyl red, found that the precip- itation of aluminum hydroxide from aluminum chloride solu- tions was complete when£hflwas lO-6°5 to 10-7'5. Joffe and McLean (8) found that in the presence of sulfate-ions the aluminum from a 0.0075 M. solution was transformed completely into the gel at pH 4.7 to 4.8. However, in the presence of chloride-ion complete aluminum precipitation occurred at pH 5.4. Prideaux and Henness (9) in studying the precipitation of hydrous aluminum oxide reported visible precipitation from the sulfate solution at about pH 4, but not until about pH 6.5 in chloride solution. While there is little agreement as to the exact pH values at which aluminum hydroxide starts to precipitate or at which it is removed from solution, the above inves- tigations are in agreement that precipitation from sulfate solution occurs at a lower pH than from the chloride. This supports the general rule that the bivalent sulfate ion has a greater coagulating effect upon a positive colloid than da:s the monOValent chloride ion. Hildebrand and Bowers (10) found that zinc was removed at a pH of about 6.3. Kolthoff and Kameda (ll) in their studies of the hydrolysis of zinc sulfate titrated various molar concentrations of zinc with sodium hydroxide. They obtained pH values of 6.17 and 6.49 at which zinc starts to precipitate from 0.05M and 0.01% zinc sulfate solution, respectively. Britten (2) reported a pH value of 5.2 for the first appearance of precipitation from a 0.025M solution. The pH values given by Britten (2) (12) for the pre- cipitation of cadmium, copper and chromium were the only ones found in the literature. The effect of certain metallic impurities upon cathode effecieney, type of deposit, etc., in nickel plating baths has received considerable attention (15) (l4) (15) (16); yet definite information regarding the pH values at which these impurities may be resoved is limited. Thompson and Thomas (24) suggested the following procedure for the removal of zinc from nickel baths; the bath is neutralized with nickel hydroxide or carbonate to pH 6.7, allowed to settle, then filtered, reacidified to pH 5.8, and electrolyzed at high current density for some hours. Blum and Hogaboom (25) discussed the effects of some metallic impurities and the pH's at which they may be removed. EVans (17) described a method of reuoving chronic acid from nickel plating solutions. In a recent publication, Pinner, Soderberg and Baker (18) have considered methods of removing impurities from nickel baths. Experimental Materials and Apparatus: - a stock solution of nickel sulfate was prepared by dissolving 2270 g. of Baker's analyzed C. P. Ni804'6U20 in 8 liters of dis- tilled water. Inasmuch as both ammonium hydroxide and so- dium hydroxide were to be used in the work to follow, this solution was divided into two portirns; one was treated with ammonium hydroxide, the other with sodium hydroxide solu- tion, approximately 0.5N, until the pH was 6.7 to 6.8. The solutions were heated for one hour at 90° C, allowed to stand twenty-fours, then filtered. The metals were added in the form of their sulfates, except in the case of hexaValent chromiun to give a metal content of approximately 0.5 g/l. Baker's Analyzed Chem- icals were use for aluminum, cadmium, ironuand nickel sulfates. he Cr2(SO4)5°5H20 and Cr05 were C. P. grades. The sulfates of copper and zinc were prepared from the pure metals. Boric acid, (Baker's C. P. grade) recrystal- lized from water was used as the buffer. Each liter of solution contained 57.5 gms of boric acid. analyses for cadmium, copper, and :irc were made employing a Leeds and Northrup Electro-Chemograph. Measrrements of pH were made with a laboratory model (G) Beckman pH meter. method of Analysis: - For the polarographic deter- mination of cadmium and copper, buffered nickel sulfate solutions served as the supporting electrolyte, but since the zinc and nickel waves come at approximately the same half-wave potential in the sulfate solution the zinc was first separated from the nickel by the method of Fales and Rare (1?), the sulfide was dissolved in cold 1:1 hydrochlor- ic acid, evaporated to dryness, taken up in 25 ml. of a solution, 0.13 ammonium acetate and 0.0253 potassium thio- cyanate, as suygested by Reed and Cummings (20). aluminum was determined by use of aurin tricarhoxy- lic acid (as). Chromium and iron were determined volumetrically using 0.02 and 0.01N potassium dichromate with diphenyl- .amine sodium sulfonute as indicator. Trivalent chromium was oxidized to the hexavalent state with ammonium per- sulfate, then the above mentioned procedure followed. Ferric iron was precipitated from the nickel sulfate solution by use of an excess of ammonium hydroxide, fil- tered, dissolved in warm 1:1 hydrochloric acid and the Zimmerman-Reinhardt method of analysis used. Ferrous iron was first oxidized to the ferric state with hydrogen per- oxide, then the hydroxide precipitated with ammonium hy- droxide. Experimental Procedure: - Five hundred ml. of the purified nickel sulfate solution was acidified to a pH of 2.0 - 5.0 by addition oi‘sulfuric acid solution. The weighed quantity of metallic salt was dissolved in the -3- solution, then three 25 m1. samples were withdrawn for analysis. ammonium hydroxide (1:4) was added dropwise while the solution was being stirred vigorously by means of a motor driven stirrer. a period of five minutes of stirring was allowed after the NH4OH additions before sa pling. at 0.1 - 0.2 pH intervals two 25 ml. samples were withdrawn and pipetted into 7" x 5/8" test tubes. One series of samples was allowed to stand at room temper- ature (2-3O - 270 C) for 24 hours with occasional shaking, while the other was heated in a boiling water bath for three hours. The test tubes in this second series were fitted with air condensers in order to minimize evaporation. These were made by almost closing off in a flame one end of a six inch piece of seven mm. tubing. at the end of these periods observations were made as to the condition of the solutions, the samples filtered through No. 40 or 42 Whatman paper when the nature of the precipitate permitted, and pH measurements again made. analyses were run on the solutions in the manner described earlier in the thesis. The above procedure was repeated using 0.2K sodium hydroxide to raise the pH. Each metal was treated in the manner previously described with the exception of ferrous iron in which case it was necessary to keep an inert atmos- phere above the solution and saoples at all times during the run in order to prevent oxidation. Nickel Carbonate and calcium hydroxide were added to a buffered nickel sulfate solution in order to determine the maximum pH values that could be obtained when the samples were thoroughly agitated at 250 C and when heated for three hours at 100° c. In the course of tFis investigation the question of obtaining the equilibrium pH values arose, since during precipitation hydroxyl ions were being removed from solu- tion, thus changing the pH. This precipitation reaction sometimes required considerable time. Accordingly, it was decided that after each addition of base five minutes be allowed for stirring before the samples were collected and their pH determined. The pH's of these Samrlrs were again measured after standing 24 hours at room temperature (220 - 280 c) in order to obtain the equilibrium values. Measure- ments of pH werealso made after the second series was heated three hours at 1000 C. These data are recorded in the tables. From a practical standpoint the pH value to which the solution must be raised in order to remove a certain amount of impurity would be most useful. The equilibrium pH values do not always give this information since they are usually from 0.2 - 0.5 pH units lower than the initial values. Thus, in plotting the graphs the initial values were recorded. Data and Discussion of Results Buffered Nickel Sulfate Solution Treated with Ammonium or Sodium Hydroxide for Three Hours at 1000.9 and 23 Hours a} Room Temperature (250 - 26° C): - In figures one to four inclusive a single set of conditions is represented on each graph. Each metal which has been studied under a given set of conditions is represented by a curve on the appropriate graph. From figures one to four inclusive the pH range in which a given metallic hydroxide will precipitate is shown and one can compare directly that range with the range in which the other metals studied come out of solution. By comparing figures one and two and figures three and four it is shown that more rapid precipitation is favored at a lower pH in the heated solutions. In the case of the ferric iron curves in figures one and two it is Shown that precipitation of ferric hydroxide begins at a lower pH in the hot solutions than in the cold ones. It may further be observed, however, that at pH 5.5 about the sane amount of iron has been re- moved in both cases. Of coirse, the case of‘heating ccn- siderahle time is eaVed. Similar canparisons may be made fcr the other metals shown in figures one through four. 7.50 7.0 I 6.0 6.0 W 3.0 >\ . \ Cr” - :5\ \ 2.0‘ 0 Fe 1.5 o 163 320 A30 Killian-e of Hotel per Liter Pig. 1 -Concentretion or eluninum, cadmium, chromium (hexavelent).ehror.- ium (trivalent), copper, ferric iron, and zinc in buffered nickel eul- rutg solution treated Iith ammonium hydroxide and heated three hours at 100 C. '1‘ ab 1e 1 Concentration of aluminum, cadmium, chromium (hexa- valent), chromium (trivalent), copper, ferric iron, and zinc in buffered nickel sulfate solution treat- ed with ammonium hydroxide and heated three hours at 130“C. metal ~ln1tial pH Final p3 lg “etal/l al (orig. soln.) ----- 492 5.42 5.15 492 5.59 5.14 497 5.55 5.18 476 5.95 5.19 445 4.07 2.75 39 4.22 2.51 53 4.53 2.62 25 4.49 2.71 21 4.72 2.82 2 4.58 2.35 19 $3 (Orig. soln.71 ----- 462 5.05 5.72 462 5.97 5.73 445 6.17 5.82 459 ‘.50 5.87 445 6.42 5.98 453 6.55 6.02 421 6.70 6.15 415 6.79 6.24 466 6.37 6.5:1 415 Hexa- (Drig. seln., ------ 455 valent 5.00 5.11 455 Cr 5.40 5.59 448 5.65 5.55 424 5.90 5.67 59) 6.03 5.30 546 6.10 5.5l 502 6.20 5.63 266 6.53 5.36 154 6.00 6.02 105 6.c8 6.25 61 6.05 6.57 42 (.00 6.51 25 Trival-(Crig.soln.) . ----- 524 ent 5.32 2.62 524 Cr 5.78 2.55 471 4.25 2.64 214 4.50 2.63 115 4.70 2.95 54 4.90 5.82 5 5.50 4.90 2 5.60 5.50 0 Table #1, C(nt'i. etul Tnitia p3 rinal 3 etal/l Cu (crib. soln.) ------ 485 1.50 4.19 235 4.60 4.12 421 1.23 4.52 415 4.97 4.55 541 5.10 4.46 301 5.27 4.55 197 5.44 4.72 95 5.68 5.18 57 5.90 5.56 44 L.00 5.63 27 6.19 5.32 9‘ 6.41 5.97 8 6.62 6.07 11 Ferric (orig. soln.) ------ 460 Iron 1.5 1.62 455 1.67 1.62 249 1.75 1.65 69 1.39 1.75 95 2.01 1.78 71 2.14 1.85 49 2.27 1.92 69 2.65 2.04 42 4.20 2.10 29 2.95 2.15 59 5.05 2.13 20 5.04 2.45 28 5.19 2.74 2 5.45 2.95 25 5.55 5.00 25 Zn (orig. soln.) ----- 452 5.55 5.45 405 5.94 5.61 244 6.05 5.65 22 6.23 .65 67 6.42 5.92 58 6.33 .59 5 6 \) Zn Cd Ci! 6.0 Cu -‘>“-~ F $3 0 5.0 o o 01""3 5+": A1 4.0 . Fe 3.0 3 20° ' O 163 320 480 Hill'grams of Letal per Liter Fig. 2 - Concentration of aluminum, cadmium, chron2un (hexavalent), chrdm- 1um (triva;ent), copper, ferric iron, and zinc in buffered nivkel sulfate solution treutod :1 th umcnium hydroxide and heated 24 hours 2:5 to 26° c. 2 :1”: Table Concentration of aluxinum, cadmium, chromium {hexa- Valent), chrofiium (trivalort), copper, ferric iron, and zinc in buffered nickel sulfatu solution treat- ed with a'wonium hydroxide and heated 24 hours at 250 to 26° c. metal Tritia? pH Tingl pi i5 notal/l n1 (orig. soln.) --- 435 5.35 5.55 435 5.95 ..95 425 4.01 4.05 435 4.12 4.15 451 4.53 .52 79 4.49 4.45 25 4.72 4.72 90 4.98 4.34 86 Cd (orig. soln.) --- 459 6.17 6.25 455 .50 6.40 415 6.42 6.42 489 6.55 6.55 555 6.70 6.66 562 6.79 6.7 559 ' 6.37 6.79 530 Hexa- (orig. soln.) --- 454 valent 5.90 5.87 454 Cr 6.20 6.20 449 6.50 6.47 425 6.35 6.35 572 7.00 7.02 552 Tri- (orig. soln.) --- 525 valent 4.55 4.00 515 Cr 4.50 4.16 561 4.70 4.46 156 4.90 4.79 77 5.50 5.25 55 5.60 5.62 54 5.75 5.78 57 Cu (orig. soln.) ~-- 485 4.80 4.62 491 4.97 4.85 494 5.10 5.02 464 5.27 5.2 452 5.55 5.28 453 5.58 5.45 554 Table metal hitial p: ina- p€ 45 ietai/i Cu 5.62 5.57 263 (con'd) 5.55 5.70 201 5.90 5.75 166 6.00 5.35 151 6.19 6.30 10 6.41 6.25 61 6.62 6.45 64 Ferric (orig. soln.) --- 460 Iron 2.8‘ 2.60 415 2.95 2.69 565 5.05 2.65 455 5.04 2.79 455 5.19 5.08 85 5.45 5.53 20 5.55 5.50 17 5.47 5.71 17 7.92 5.88 15 Ln (or g. soln.) --- 452 6.17 6.29 477 6.52 6.67 275 6.88 7.01 175 600 ‘ 5.0 db k 4.0 D f (D 0, CI’ 5 0 3.0 \ 0 200 x .Fe 1.0 O 160 320 480 Milligrams of ketul per Liter Fig. 3 - Concentration of aluminum, chromium (trivalent), and ferric iron in buffered nickel sulfate solution treated with sodium hydroxide and heated three hours at 106° c. Table #5 Concentration of aluminum, chromium (trivalent) and ferric iron in buffered nickel sulfate solution treated with sodium hydroxiie and heatej three hours at 10000. acnmvb 5‘3 01 C21 '51 CH 1'“ OQOCJQCH' 00 3.13 5.17 3.13 metal Initial pr Final p2 - kg Ietal/l Al (or g soln.) --- 472 o 5012 4WD 0 3019 475 415 438 445 1.5* 1.72 1.88 2.08 2.22 2.50 2.70 2.82 2.98 5.08 5.06 5.15 5.45 5.82 1.61 1.62 1.71 1.81 1.28 2.02 2.11 2.15 2.19 9 0'7 H.111. 2036 2.69 2.95 5.15 4.09 2.94 107 4.20 2.62 77 4.28 2.5 44 4.42 2.52 50 4.65 2.7 24 4.85 2.80 24 Trival- (orig. soln) --- 472 ent Cr 2.89 2.55 472 5.15 2.53 59 5.40 2.53 284 5.62 2.61 259 5.82 2.61 254 4.00 2.59 194 4.22 2.59 150 4.53 2.51 112 4.58 2.69 55 4.74 2.91 19 4.87 5.17 9 5.02 5.72 2 5.50 4.88 0 Ferric (orig. soln.) --- Iron 1.43 1.5 UMFCHUI 5L 0: 01 (fl Fu54-¢ cncsq (DGJO.\ Cribb- LD r103 OHco' I0 {‘0 C. " c p.» (ECO 12 7.0 6.0 1 A1 + 4.0 '0 Fe 3.0 - _ 10“ \ o o 8.6 o 160 520 430 milligrams of metal per Liter F13. - 4 Concentration of aluminum, chrorium(tr1velent), and ferric iron in buffered nickel sulfate solutgon treated with sodium hydroxide and heated 24 hours at 23° C. #4 I.‘.1"1,’> §‘—\’ Concentration of aluminum, ch omium (trivalent),and ferric iron in buffered nickel sulfot- solution treated with soiium hydroxide and heats? twenty four hours at 250 C. Metgl :q.t:A1 p" riqul pH .r ietai/i It]. ((1514. 301:1.) --- 4,72 4.00 5.73 472 lqu 4006 472 ‘ 2“ 4.15 472 4 27 4.72 550 4.42 4.57 73 4.65 4.65 7? 4.35 4.36 75 5.10 5.23 C- Trival- (orig. soln.) --- 468 out Cr 4.58 4.02 449 4.5 4.17 540 4.52 4.27 255 4.74 4.43 120 4.87 4.67 71 5.02 4.97 2% 5.50 5.23 16 5.42 5.45 4 Ferric (OT18.SOln.) -‘“ 455 Iron 2.45 2.5~ 455 ..2.70 2.74 419 2.82 2.85 422 2.93 2.70 292 5.04 2.68 563 5.06 2.75 550 5.03 2.85 408 5.15 5.01 72 5.45 5.5 14 5.82 5.65 20 m-—— a -- v Treatment with Nickel Carbrnate, Calcium hydroxide, and Magnesium oxide: - Since nickel carbonate and calcium hy- droxide are often used in purification of nickel plating baths it was thought desirable to determ‘ne the maximum pH values that could be obtained with b.ffered nickel sulfate solutions. Tables a and B show the results of this study. The data in table A were obtained upon treating buffered nickel sulfate solutions with nickel Carbonate, calcium hydroxide and magnesium oxide and allowing them to stand at 220 - 270 C with intermittent shaking. It will be seen that in the buffered nickel sulfate solution nickel carbonate failed to raise the pH to 6.0. Calcium hydroxide did raise the pH above 6.0 in the unheated samples but only to pH 5.4 in the heated samples. The data indicate very definitely that the proper procedure to attain the maximum pH is to allow the nickel carbonate or calcium hydroxide to react with the buffered solution at roan tem- perature. This is in accordance with the fact that the sol- ubilities (27) of calcium hydroxide and nickel carbonate are much less at the higher temperatures. The heated series in table B was carried out by adding the nickel carbonate and the calcium hydroxide at room temperature, then heating to 100° C for three hours. When the solutions were heated to 1000 C, the nickel carbonate and the calcium hydroxide added and the heating continued for three hours, the data in table B were obtained. -11- Table A pH Values of Buffered NiCkel Sulfate Solutions Treated with 5 g/l of Nickel Carbonate and Calcium Hydroxide. Solution Reanent In'tl pH Time 5 Temp. Final pH Buffered N1304 NiC05 2.70 5 hrs. at 25C 0 5.90 Buffered N1504 x100 2.70 5 hrs. at 100° c 5.22 Buffered NiSC Ca(oh)2 2.70 5 hrs. at 25° 0 .50 4 Fuffered N150. CaCOH)2 2.70 3 hrs- at 1000 C 5°40 ‘1 Table B This table gives the data collected when the solutions were heated to 1000 C, then the reagent added and the heating continued for three hours at 1000 C. Solution Reagent Initial pH Final pH Buffered NiSO4c NiC05 5.00 5.12 Buffered N1504 111005 5.04 5.52 Qualitative Effects of the Use of Urea in Raising the 2H 2; Nickel Sulfate Solutions: - Willard and Tang (28) and Willard and F053 (29) have carried out extensive investi- gations in the use of urea as an internal method of rais- ing the pH of an aqueous solution. Urea in solution, upon being heated hydrolyses to give ammonium hydroxide and carbon dioxide. If the urea is added to the cold solution which is then vigorously shaken and heated the pH of‘that solution will slowly and homogeneously be raised. A very brief exploratory study was made in order to observe the effects of urea on the pH of a concentrated nickel sulfate solution. No toric acid was used in this case. The pH of the nickel sulfate solution was adjusted to 1.5 with sulphuric acid. Small quantities of the urea gradually increased the pH. Further addition of‘the salt brought about precipitation. A pH of 6.6 was reached. Further additions of urea produced no change in pH until the nickel concentration was 5reatly reduced, the nickel hydroxide precipitating at pH 6.2 ~ 6.7. Some qualitative runs were made with certain metallic ions as impurities (chiefly Fe***). The precipitate formed was easily removed on the filter, leaving the nickel sulfate solution quite clear. Additional tests in the presence of the chloride ion indicated that the results were the same. I '...J O] I The chief advantage in the use of urea is that the pH is raised slowly and homogeneously throughout the entire solution, thus avoiding abnormal pH changes in the vicinity of added Trecipitant. L Discussion of the Individual Ions Studied aluminum: - Initial D? rpenrgnce if tughidiff*iwvn1 addition of either annonium or sodium hydroxide to buffered unheated nickel sulfate solution containing aluminum sulfate was osse-ved at pH 4.4 i 0.1, although buff r action.was encoun- tered at pH 4.0. The solution rapidly became very turbid upon further addition of base and at pH 4.6 - 4.7 coagulation of the hydrous aluminum oxid: was noted. as seen from fi' ure 5, the concentration of the aluminum was markedly A ‘ ..4 fl C) (‘ reduced at pH 4.4 upon standin- for 24 hours at 2 Figure 5 further shows tlat heating tne saxples three hours L); O . at 100 C produce heavy precipitation at 0.5 - 0.4 pd units lower, however, the aluminum was not completely removed at pH 4.9. It may also he obServed that in the T case of the heated suuples a considerable drop in p? occur- red even where no visible precipitation took place, indica- ting hydrolysis. Some difficulty was encountered in the use of aurin tricarboxylic acid for low concentrations of aluminum. In order to have sufficient aluminum (0.1 - 0.5 mg.) for the analysis 20 ml. sanples were taken. The high nickel sulfate content caused coagulation and settling of the lake so that it was necessary to shake the colorimetor tufies frequently while making check readings. 5N 4.90 4.50 \‘3 4.10 \ .—— s.soL i ‘ l o 150 320 480 Milligrsns per filter of Aluminum Fig. 5 - Aluminum in buffered nickel sulfate solution at different pH values: 1, uses treatment, 3 hrs at 10000.; 2, on treatment, 3 hrs at 10000.; 3 uses treatment, 84 hrs at 2590.; , NH4OH trest- nont. 24 hrs at 2560. -15.. Cadmium: ~ It may be noted in figure 6 that cadmium is not appreciably removed from a buffered nickel sulfate solution even at a pH of 6.6. In fact it appears that some cadmium returns to the solution above pH 6.7. It was found that in a boric acid solution cadmium is almost entirely removed from solution at pH 7.$ while in a buffered nickel sulfate solution only 60 mg/l of cadmium was removed from solution after it was heated three hours at 1030 0. While making the run with cadmium in the buffered nickel sulfate solution it was obserde that between pH 6.2 and 6.5 the solution became very turbid and above 6.3 a laroe quantity of nickel hydroxide was precioitated. This was about 0.6 pH unit lower than the value at which nickel hydroxide started to preci;itate in an unbuffered solution (51). A run was made in which the pH of a buffered nickel sulfate solut on was raised with ammonium hydroxide. The results are shown in figure 7. In View of the heavy precipitation of nickel hydroxide above pH 6.3, the removal of cadmium from the buffered nickel sulfate solution cannot be attributed to the influence of the boric acid since it is entirely probable that some of the cadmium was adsorbed from solution by the nickel hydrox- ide. It is evident also that any visual observations as to *3 the pH values at which metallic hydroxides come down. e worthless if these values fall within the range where pre- cipitation of nickel hydroxide occurs. In this range a -15- chemical analysis must be made to determine the cadmium present. Only ammonium hydroxide was used in the study of cad- mium due to the dilution factor that would have been en- countered had sufficient dilute sodium hydroxide been used to reach the desired pH. 7.00 6.80 6.60' 6.40 6.20 6.00l I 240 320 400 ' 430 Milligrams per Liter of Cadmium F1 . 6 - Cadmium in buffered nickel su fate solutions at different pH values: 1, Buffered nickel sulfate solution after NH4OH treatment, 3 hours at 106° C. 2, Buffered nickel sulfate solution after NH40H treatment, 24 hours at 23°C. 7.00 6.60 6.20 5.80 5.40 5.00 20 Grams par Liter of Nickel 40 60 F13. 7’- Nickel content of buffered nickel sulfate solution at differ- l, NH4OB treatment, 3 hrs at 10000.; 2, NH4OH treatment out pH values: 24 hrs at 25°C. -17- Hexavalent Chromium: - A buffered nickel sulfate solution containing 500 mg/l chromium.was used. The pH was raised With ammonium hydroxide. As may be seen in figure 8, the break in the curve occurs at pH 5.4 in the case of the heated sample and 5.8 in the cold sample. As shown in figure 8, a pH of 7.0 is required to remove chromium almost completely from solution. It was ftund that the precipitation from buffered nickel sulfate solutions with ammonium hydroxide (upon heating) is more efficient than in unbuffered solutions. A curve for the unbuffered precip- itation approaches that of the buffered solutions above pH 6.8. The voluminous precipitate of nickel hydroxide in the buffered solution probably carries down some chrom- ium with it. It seems doubtful that the boric acid has any influence in this case since it was shown that hexa- valent chromium does not precipitate at all from a boric acid solution up to a pH 7.1 when armonium hydroxide is used as the precipitant. With sodium hydroxide used as the precipitant, a small quantity of light yellow needles was observed at pH 7.0 after the solution had been heated three hours at 103° c. The solution changed progressively from an orange-red at pH 5 to a light yellow as sodium hydroxide was added, in- dicating the conversion of the chromic acid to sodium chromate. o i 6.60 l i : l l 6.20 . '1 pii I j z I I I 5.80 = k 0 5.40 1 5.00 o 160 320 480 6 milligrams per Liter,of Chromium Fig. 8 - Hexevelent chromium in buffered nickel sulfate solution at different pH values: 1, NH4OH treatment, 3 hrs at 10000.; 2, NH4OH treatment, 24 hrs at 24°C. _ -13- Trivalent Chromium: - In shidying trivalent chromium in the buffered nickel sulfate solutions it was found that initial turbidity, in the case of the heated sa plea, occurred at pH 2.8 - 2.9 when sodium hydroxide was used and at pH 3.4 - 3.7 when ammonium hydroxide was employed. In the case of the unheated samples turbidity was first noticed at pH 3.4 - 3.5 for both hydroxides. To obtain reproducible data it is necessary to add the hydroxide slowly and with constant agitation. It is quite possible that rapid addition of the base could cause localized changes in pH sue? that nickel hydroxide would be precip- itated, thereby causing turbidity due to nickel hydroxide instead of chr;mium hydroxide. Of course, this same line of reasoning may be applied to all the metals which start precipitating in pH ranges below the pH at Which nickel hydroxide comes out. As shown by the curves on figjre 9, heating the solu- tions increases their acidity, i. e., precipitation begins at a lower pI-I value. In the case of the samoles heated at 1000 C, there was considerable difficulty in obtaining consistent data concerning the minimum pE point at which precipitation oecurs after three hours. after five trials using anmonium hydroxide as the precipitant, it was decided that the precipitation probably begins at some pH under 2.0. It is probable that the inconsistent results is due to the rate of hydrwlysis of the chronic sulfate. Britton and fiescctt (12) carried out titrations of 0.5 h uhrrmic salts with 2.0W sodium hydroxide. They observed an increase in acidity upon heating the salt sol- utien as shown by the fact had a pH of 2.85, but when the pH had dropped to 1.2. that a 0.3 M CrClS°CH20 solution heated to boiling and cooled 6.80 6.00 ; i . L i 5.20 i t‘—‘ ‘ “\K\\\\\.\\‘ \ 3 .1 z hl“_ -\\\\\\<§ 3.60 \\L ‘ \ 2‘804 140 Stfi Milligrams per Liter of Chromium ria.€3 - Trivelent chromium in buffered nickel sulfate solution at different pH values: 1 treatment, 3 hrs at 100 NeOB treatment, 3 hrs at 10000.; 2 .; 3, secs treatment, 24 hrs at 23 33403 treatment, 24 hrs at 23°C. be??? -23- Copper: - In the treatment of copper only ammonium hydrox- ide was used. In solutions whose pH was 5.6 - 5.7 immediate- ly after the addition of ammonium hydroxide, turbidity was observed in the unheated sample at pH 5.0 - 5.1 and in the heated salple below 4.3. Figure 10 shows quite clearly that copper is nore nearly removed from solution in the heated sample. at pH of 6.6, ll ng/l of copper remains in the heated solution while 64 nq/l remains in the cold sample. Copper is removed quite rapidly from a boric acid solu— tion at pH 5.4. Reca_ling that boric acid is used as a buffer it is understandable why the copper behaves as it does in this case. 6.60 6.20 5.8 ‘\\\\ \‘¢\ \ 0 \° 5.00 \\g 11 l ‘0 l > 0 4.60 ‘——» ' 520 480 160 Milligrams per liter of Copper Fig. 10 - Copper in buffered nickel sulfate solutions at different pH values: 1, Buffered nickel sulfate sclution after NH OE, treat- ment, three hours at 10000.; 2, Buffeged nickel sulfatetsolution after NH4OH treatment, 24 hours at 24 C. Ferric iron: - The remOVll of iron from a buffered nickel sulfate solution is of great interest, especially to tte electroplating industry. As s‘own,on.figure 11, the pre- cipitation of iron from a hot solution begins below pH 1.5 upon the addition of either sodium or ammonium hydroxide. The behavior of a cold solution is quite different. Turbidity becomes evident at pH 2.8 - 2.9 when sodium hydroxide is used and at 2.6 - 2.7 when ammonium hydr xide is added. Upon heating a series at 1000 C coaéulation and volumin- ous precipitation was observed in all samples above pH 1.0 after five minutes in the water bath. Over the period of three hours the sanples below this pH showed precipitation down to pH 1.6 - 1.7. Between pH 1.5 and 1.6 the solutions were turbid, but no settled precipitate was present. At pH 1.4 the sa ple was clear. ann figure 11 it may also be noted that in the case of the unheated samples there is a very narrow pH range (5.0 - 3.1) in whids the colloidal ferric hydroxide is quite stable. This stability was found to persist for 36 to 48 hours; however, at pH near 3.5 the iron content in all cases is reduced by 803. In the case of the heated samples it is lowered by about 803 at a pH as low as 2.0 - 2.1. The fact that iron is so efficiently removed at such low pH values provides a very convenient method of clearing a Watts Nickel plating bath of iron impurities. 3.50 D O 3.10 G ! I T Q. -___-~‘ 2.70 O 3 4, PH Or ‘\b' 2.30 1.90 \ 1.50 O 160 320 480 Milligrams per Liter of Iron Fig. 11 - Ferric iron in buffered nickel sulfate solution at different pH values: 1, NaOH treatment, 3 hrs at 10000.; , NH4OH treatment, 3 hrs at 10000.; 3 NaOH treatment, 24 hrs at as 0.; 4, NH4OH treat- ment, 24 hrs at 236C. Zinc: - Zinc was not appreciably removed from the buffered nickel sulfate solution with an onium hydroxide until a pH of 6.4 was reached as shown in figure 12. At about pH 6.4 the curves break quite rapidly. No sodirm hydroxide was used due to a dilution factor, therefore, the two curves shown on figure 12 are for the use of ammonium hydroxide at 1000 C and at 250 c. It may be noted that the zinc is removed completely from solution at pH 6.8 in the heated solution. . In this partiallar case the advantage of the heated process over the cold process is especially obvious. Again in this case, visual observations were of no value. analysis showed that in the heated sample zinc was precipitated at a much lower pH in this solution than in nickel sulfate solution alone. The possible effects of the precipitation of nickel hydroxide must not be overlooked. In Carrying out tte analysis, which was a combination of the separation method of Fales and flare (l9) and the polaragraphic procedure of Reed and Cum ings (20), several points of importance should be noted: (a) In order to have sufficient zinc to work with conveniently at the lover concentrations fifty m1. samples were taken; (b) In the formic mixture, the use of either ammonium chloride or ammonium sulfate Was recommended, for this case ammonium chloride was found preferable, since there was a large amount of nickel ammonium sulfate precipitated otherwise, as well as some zinc aflmonium sulfate; (0) It was essential that the solution be at pH 2.0 before saturation with hydrogen sulfide gas, otherwise nickel sulfide orecipitation was incomplete. A small auount of nickel sulfide is not obiect- ionable since the zinc sulfide may be dissolved in cold 1:1 H01, leaving the nickel sulfide. 7.00 6.60 6020 \ 5.80 5.40 o . 150 ' 320 430 Milligrams per Liter of Zinc Fig. 12 - Zinc in buffered nickel sulfate solution at different pH values: 1, Buffered nickel sulfate solution after NH4OH treat- ment, 3 hours at 100°C. 2, Buffered nickel sulfate solution after NH on treatment, 24 hours at 25°C. General Conclusions: - In general the removal of impurities from buffered nickel sulfate starts at a lower pH in heated solutions than in the cold ones. As might be expected, more efficient r moval of the impurities was effeoted in the solutions that were heated for three hours at 1000 C, t1an_in those that were allowed to stand at 220 - 270 for 24 hours. In the case of aluminum, trivalent diromiun, copper and ferric iron the hydrolysis process manifest itself in the heated samples in the form of a drop in pH amounting to 0.5 - 2.5 units. Summary: - The removal of aluminum, cadmium, trivalent and nexavalent chromium, copper, ferrous and ferric iron and zinc from buffered nickel sulfate has been studied at 220 - 270 C and at 1000 C usin; ammonium and sodium hydroxide whenever possible. The maximum pH values attained when buffered nickel sulfate solutions were treated with 5 g/l of nickel car- bonate and calcium hydroxide and thoroughly stirred for six hours at 220 - 270 C or heated for three hours at 1000 C were obtained. Urea was used as an internal agent for the adjustment of pH. It was found that upon heating a nickel sulfate solution to which urea had been added, the pH increased to 6.6 at which point nickel hydroxide separated out. Literature Cited Hildebrand, J., J. mm. Chem. Soc. §§, 847 (1915) Britten, H. .s., J. Chem. Soc. 121, 2110-2156 (1925) Atkins, Trans. Faraday Soc. lg, 510 (1925) Patten and mains, J. assoc. Off. agr. Chem. 3, 233 (1920) Pickering, S.U., J. Chem. B00. 31, 1991 (1907) Miller, L.B., U.S. Public Health Reports EB, 1995 (19 25) Blum, 3., J. Am. Chem. Soc. 53, 1232 (1916) Joffe, J.S. and McLean, H.C., Soil Science_§§, 47-59 (1928) Prideaux, E.B.R. and Heness, J.R., Analyst fig, 85 (1940) Hildebrand, J. and Bowers, J. am. Chem. Soc. 38, 785 (1916) ‘— fiolthoff, 1.x. and Kameda, T., J. Am. Chem. Soc. 2;, 852 (1951) Britten, H.T.S. and Wescott, C.B., Trans Faraday Soc. 22, 809 (1951) JacNaughton, D.J. and Hammond, Trans. Faraday Soc. fig, 431 (1950) Thomas and Blum, Trans Electrochem. Soc. i9, 69 (1925) hadsen, Trans. Electrochem. Soc. éé, 249 (1924) Haring, Trans. Electrochem. Soc._g§, 107 (1924) Evans, B.S., Analyst fig, 58 (1921) Pinner, W., Soderberg, G. and Baker, E.L., Trans. Electrochem. Soc. 22, 554-574 (1941) Fales, H.A. and Ware, G.H., J. Am. Chem. Soc. 21, 487 (1919) Reed, J.F. and Cumnings, R.W., Ind. and Eng. Chem., anal. Ld.‘l§, 489 (1940) 26. 27. 28. 29. 59. 51. Literature Cited (Con't) A! Lamb, n.B. and Fonda, u.H., J. am. Chem. Soc. 45, 1:54 (1921) monte - Lartini, C. and Vernazza, E., Industria Chimica Z, 857 and 1001 (1952) Denham, 2., anorg. Chem. 53, 531 (1908) Thormlson, L.R. and Thomas, C. T., Trans. Am. Electro- chem. Soc. 45, 79 (1922) Blum and Hogaboom, "Pri1ciples of Electroplating and Electroformin5" p5 255-255 (1950) Snell, "Colorimetric Methods of Analysis" Vol. 1, (1956) Seidell, "Solubilities of Inorganic anc Or5anic Com- pounds" Vol.1 Lillard and Tang, Ind. Eng. Chem. Anal. Ed. 9, 557 (1957) ‘ Willard and F053, J. Am. Chem. Soc. 59, 2422 (1957) Hillard and Tang, ibid, 59, 1190 (1957) Biesner, H. J., (Thesis iichi5an State College 1945) "The Colloidal State and Precipitation of Certain metal- lic Hydroxides in Concentrated Solutions of Nickel Sulfate". MY 2. 4_ '01 T542 . 6 217618 11128 E-Ic01uggage \\\\\\\\\\\\\\\\\\\\\ M); 7817 \ 2 I ‘III (“I‘ll “H \ I“) \IIIIII A \ Tl \ S I“ III‘ MI)“ MW 293 O (1'