Dowex 50 - Industrial & Engineering Chemistry (ACS Publications)


Dowex 50 - Industrial & Engineering Chemistry (ACS Publications)pubs.acs.org/doi/abs/10.1021/ie50464a003by WC Bauman...

32 downloads 86 Views 2MB Size

S'IYMPOSI'IIJMON

Purification and Conditioning of Water Supplies

I

Presented before the Division of Water, Sewage, and SI

DOWEX 50 A NEW HIGH CAPACITY CATION EXCHANGE RESIN W. C. Bauman, J. R. Skidmore, and R. H. Osmun The D m Chemical Company, Midland, Mich.

High capacity and excellent chemical stability are the chief attributes of Dower 50, a new cation exchange resin of the sulfonated h y h r b o n type, which is now commerc i d y avdable. In water softening and demineralization it hm shown operating capacitiea nearly double that of meins previously availnble. The e l i m i n a t i o n of the phcmolic p u p fmm the -in structure greatly enhanees the chemical stability and permits operating i n a wider r a n g e of pA, in the presence of oxidizing agents, a n d in watrre up M 1000

c.

agents. The phenolic group, however, lends instability to the resin structure at high pH vduea and under oxidizing conditions; it lahilieeg some of the sulfonic acid groups at high temperatures in the acid form. These effects are shown in Figures 1and 2 4s recent data run on a cation exchanger of this type (8). The subject of this paper is a new cation exchange resin, Dowex 50, with a cross-linked aromatic hydrocarbon c h i n coutginiug nuclear sulfonic acid groups as the sole cation active group (4). The elimination of the phenolic group ha8 enhanced the stability of the structure and the sulfonic acid

C

ATION exchange resina have been syntheaieed with a variety of cationic groups-that is, -ON&, -COONa, S O I N a , -CHS08Na, etc. (6). The sulfonic acid resina have been used most extensively in the general field of ion exchange because the high acid strength of the sulfonic acid group furnished a constant exchange capacity over a broad p H range in a neutral cycle and excellent conversion of salts of strong or weak acids to the comesponding free acids in the acid cycle. Exchange materials have been produced for many years by the sulfonation of n a t d materials such as coal, lignite, and peat. Them products, in general, contain both sulfonic and carboxylic acid groups b e c a w of an oxidation reaction during sulfonation. More recently owing to the pioneering work of Adam and Holmes (I), sulfonic acid resins of the phenol-formaldehyde type have been manufactured and used Bxtensively. Thew resina have p r o d very stable chemically and physically in the neutral and acid p H ranges at room tcrnpmture and even at elevated temperatures m the absence of dissolved oxygen or other oxidiew

1350

i

9 g

~

3

w,

?*

;

.o

E ?' ~

6

E

,, 2,

zo ~

I~

,o

,D

,,YE

Fi-

~

IN

,~ ,~

~

DAIS

1. Stability of Dowex 30 and Dowex 50 a t 95' C.

INDUSTRIAL AND ENGINEERING CHEMISTRY

August 1948

0

20

40

60

80

100

110

140

180

160

DECREES CENTIGRADE

200

L

Figure 2. Stability of Dowex 30 and Dowex 50 i n 16-Hour Treatment

groups, and has permitted an increase in the volume exchange capacity to 52.8 kilograins of calcium carbonate per cubic foot with maintenance of sufficient cross linking for physical stability. A previous paper (3) discussed the ion exchange equilibrium and rate characteristics of Dowex 50. The present investigation is directed toward the determination of the operating characteristics of this new exchange material in the softening and demineralization of water. The chemical analysis of a representative sample of this product a s shipped in the wet sodium form is shown in Table I. A typical titration of the acid form in the presence of 0.01 N sodium chloride is shown in Figure 3. The titrated capacity of 2.4 me. per ml. of resin is equal t o 52.8 kilograins of calcium carbonate per cubic foot and corresponds t o a sulfur content (occurring as -S03Na) of 8.5% by weight, a value in substantial agreement with the sulfur analysis.

Table I.

1351

change capacity and more uniform flow distribution. It provides fewer points of fracture with greater resistance t o attrition or degradation than any other particle shape. Spherical particles have a higher settling rate and a lower bed expansion than irregular particles of comparable size. The pressure drop in downflow operation, however, is greater for spherisal particles (Figure 16) and increases more rapidly in turbid waters because of better filtration characteristics. I n addition the exchange velocity is somewhat slower for spherical particles than for irregular particles of ground resin of equal size. For research work on special exchange problems experimental quantities of Dowex 50 have been made in smaller sphere sizes. The product has been made in the form of perfect spheres in any mesh size smaller than 16 meshes per inch down to colloidal spheres of about 0.04 micron in diameter. Spherical particles larger than 16 mesh are not generally physically stable. By the screening of various experimental batches, i t is possible t o obtain material of very uniform sphere size. Indeed, if one takes .those spheres which lodge in the meshes of a screen, one obtains almost identirally duplicate particles. Such material should be of particular interest in fundamental exchange rate studies. CHEMICAL STABILITY

In Figure 1 is shown the chemical stability of Dowex 30, a phenolic resin, and Dowex 50 in both sodium and hydrogen forms to aqueous solutions maintained at 95 C. and of p H values from 0 t o 11. Dowex 30 in the sodium form is fairly stable only at 7 to 10 pH, whereas complete stability at all p H values of both sodium and hydrogen forms of Dowex 50 is observed at the end of the 85-day test. At higher temperatures, as seen in Figure 2, Dowex 50 is again notably more stable than Doffex 30. Comp!ete stability of all forms of Dowex 50 in alkali, acid, or neutral water is noted a t 150" C. for 16 hours; the sodium form is stable in water at 175DC. for the same length of time.

Chemical Analysis of Dowex 50 i n Wet Sodium Form as Shipped Carbon Hydrogen Sulfur Sodium Oxygen ( b y difference) Water

Per Cent 26.8

7.0

8 R 0.0 51.3

41.4.

PHYSICAL FORM

This resin is available in commercial quantities in the form of spherical granules (Figure 4),the particle size of which is shown in Table 11.

Table 11. Wet Mesh Analysis of Dowex 50 in Sodium Form as Shipped On On On On On

20-mesh 30-mesh 40-mesh 50-mesh pan

Per Cent 50.0 35.0 12.0 2.5 0.5

The true density is 1.41 for the wetted spheres and 1.55 for spheres dried a t 100 O C. A cubic foot of operating bed will contain 55 pounds of wet sodium resin and about 35% of free space. The expansion of a bed of this resin on backwashing with Midland, Mich., city water at 25 C. increases as follows: 7.1 % at 3 gallons per square foot per minute; 11.5% at 4 gallons; and 16.9% at 5 gallons. The spherical form permits more uniform and closer packing in the exchange bed than would be obtained with ground, irregularly shaped resin particles; thus, it furnishes a higher volume ex-

PH

Figure 3. Glass Electrode Titration Curve of 7 Grams of Wet Dowex 50 H-Resin i n 100 Cc. of 0.01 N Sodium Chloride Solution

The common oxidizing agents to which ion exchangers are generally exposed are dissolved oxygen and chlorine. Chlorine, generally present in small amounts, has been quantitatively removed by a 16-inch bed of Dowex 30 from room temperature water containing 100 p.p.m. of chlorine at 5 gallons per square foot per minute. The same depth of bed after 6 hours of flow at 5 gallons per square foot per minute yielded a n effluent of 27 p.p.m. of chlorine from 340 p.p.m. of chlorine influent. This resin

I

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

1352

Figure 4

lost ISTOof its original volume ion %.;change capacity and had an increased water holding capacity (per cent moisture of drained dry resin) of 17.570 above its original water holding capacity. I n a similar test with Dowex 50 a 188 p.p.m. of chlorine influent water

was reduced only to 50 p.p.m. a t the start; after only 0.5 hour the effluent contained 70 p.p.m. of chlorine. The chlorine adsorption on or in Dovex 50 does not affect either its water holding capacity or its volume ion exchange capacity. Field tests show that chlorine of a 0.2 p.p.m. of chlorine influent is quantitatively absorbed by Dowex 30 but is unchanged in passing through Dov ex 50. To study the degradation of the resins by dissolved oxygen, an accelerated test was set up in which Midland city water a t a pH of 9.5 t o 10 with a chlorine residual of 0.4 p.p.m. was heated from l 5 O C. to between 60 and 70" C. and passed (along with the evolved air) upflow through 21-inch beds of the two resins. *A. system mas provided to allow the air bubbles to work up through the resin beds and the flow adjusted to give 25% bed expansion. At the end 'of the first neek a n appreciable change in the volume of the D o n e s 30 wis noted, and at the end of 3 weeks it was throwing considerable color. The water holding capacity had doubled, and its volume ion exchange capacity had been reduced 50%. At the end of 7 weeks the Dowex 50 had retained its original ion exchange and water holding capacities. It is felt that this failure of Don ex 30 is due to dissolved oxygen, as the amount of chlorine is far below that observed, t o produce such changes in the resin. Also, the resin was previously shown t o be stable t o pH 10 at higher temperatuies for a longer period of time in the absence of chlorine and dissolved oxygen. , PHYSICAL STABILITY

The water holding capacity of Dowex 50, as of other resinous exchangers, is dependent on the salt or electrolyte concentration. I n downflow operation with high concentrations of regenerant this becomes important, as the bed shrinks when i t is regenerated and retains this reduced volume until it is backwashed. Under such conditions the bed is under pressure in the softening cycle; the amount of such pressure is dependent on the concentration of the regenerant and the electrolyte content of the water being softened, I n the range of concentrations recommended for the regeneration of Dowex 50, Figure 5 shows that the degree of bed shrinkage is a direct function of the molar concentration of the regenerant. The percentage shrinking shovn is t h a t of the SO-

HESO* R E G E N E R A N T

V 0

Vol. 40, No. 8

1

I 1

2

3

4

5

6

7

MOLESlLITER

Figure 5. Shrinking of Dowex 30 in the Various Regenerants

2 x

25

1

I

15

10

5

I 20

25

30

N A C L C O N C E N T R A T I O N I N W T PERCENT R E G E N E R A N T DOSAGE 0 10 L8S N o C L I C U F T A 15 L B S N a O L I C U F T rn 20 L B S NoCL/CU FT

Figure 7 . Effect of Regenerant Concentration on Capacity of Dowex 50 i n Softening 30-Grain Water

i

I

*

20

1

I5

I

I

I

I

30

45

60

75

I

90

I

I

103

I20

HARDNESS IN GRAINS C A C O 3 I G A L T

REGENERATION. e l o LBS. 20% HoCL/CU.FI.

I 2

3 (CA"

4

A

I

1 i

5

6

7

15 L B S .

20% N O C L / C U . F t .

2 0 LBS. 2 0 % NoCL/Cu.FI. 0

9

EP / C A T I O N EO ) IN SOLUTION

Figure 6. Dowex 50 Calcium-Sodium Equilibrium

IO

Figure 8.

Effect of Hardness on Capacity of Dowex 50 i n Na + Cycle

August 1948

INDUSTRIAL AND ENGINEERING CHEMISTRY

1353

which the ordinate represents the ratio of calcium ion equivalents on the resin to the total cation equivalents on the resin. The abscissa represents the ratio of calcium ion equivalents in solution to the total cation equivalents in the solution The resin is highly selective for calcium in fresh water reducing-for example, 10 grains of hardness to 0.4 grain a t 90% exhaustion of the resin in a batchwise reaction. On the other hand, efficient regeneration requires passing the regenerant solution (in this case salt solution) through a bed of the exchanger or some other suitable countercurrent system.

10

I I

LBS. N A C L

Figure 9.

1-

/KILOGRAIN CAC03

Salt Efficiency

dium in salt solutions and that of the hydrogen resin in acid solutions. The hydrogen resin occupies 5% more volume than the sodium resin in distilled water, This shrinkage during the regeneration and compression during the softening cycle cause first, a uniformly increasing pressure drop through the exchanger bed in the softening cycle; and secondly, a very slight though measurable attrition or breakdown of the resin particles. For example, after 3000 cycles from saturated salt to distilled water, the percentage of broken spheres changed from 7 to l2y0. This does not mean a loss of material but merely the cracking of 570of the spheres (no doubt a small amount of fine material was produced, but it was immeasurable). The effect of this treatment on other resins will be studied as spherical materials become available. This resin, when given the recommended attrition tests (6) had a n insignificant increase in relative surface and no increase in per cent of broken spheres (more sensitive than the recommended test). The percentage of broken spheres is readily determined by first, drying the resin in air to free-flowing; secondly, allowing the spheres of a measured sample to roll off an inclined plane leaving the broken spheres behind; thirdly, measuring the broken material; and fourth, calculating the per cent broken. ION EXCHANGE PROPERTIES

The ion exchange characteristics of Dowex 50 in the softening cycle are perhaps best demonstrated in the equilibrium data for the sodium-calcium exchange. The several equilibriums were reached by: placing different amounts of sodium resin in calcium chloride solutions; and placing different amounts of calcium resin in sodium chloride solutions, providing agitation and permitting sufficient lapse of time to reach equilibrium. The analysis of the resultant solutione provides the data for Figure 6 in

. L:

2

45 I

,

,

0

0" 2 3s v1

5

4

a

25

0

3 5

The actual operating conditions of the resin have been evaluated both in field tests and in laboratory tests in accordance with the recommended procedures except as noted. For the tower tests herein reported, the volume of the bed is assumed t o be constant far the various regenerating conditions and was measured after: (A) thorough backwashing the sodium resin with disso I I . I I I tilled water; (B) allowing the bed t o settle for 0.5 hour; 40 and (C) passing distilled water downflow for 2 hours at 5 gal302 FLOW I4N G A L S I ~ P . F T /MINUTE lons per square foot per minute. This produces a bed volume REGENERATION almost identical with 0 10 L B S 2 0 % N C C L A. 15 the standard test pro1 2 0 " " " cedure with a spheriFigure 11. Effect of Flow Rate cal resin, and gives o n Capacity of Dowex 50 i n a more reproducible Softening 30-Grain Water figure with ground resins than does the standard test. The laboratory tower tests were run in both 1inch and 2-inch beds of 30-inch depth as outlined by the American Water Works Association (6) and in s u g g e s t e d s IO I5 20 25 30 changes recomn c L CONCENTRATION IN W T P E R C E N T mended at the REQENERATION J a n u a r y 1947 6 LBS HCLICUFX meeting of the A 9 LBS H C V C U F Z 0 12 LBS H C L I C U FT. C o m m i t t e e on Specifications for Figure 12. Effect of Hydrochloric and Methods of Acid Regeneration on Capacity of Testing Zeolites. Dowex 50 Operating in 30-Grain Except where Water otherwise wecified, the water used was made up by adding weighed amounts of reagent grade calcium chloride and magnesium chloride t o distilled water; there were 2 grains of the calcium salt as calcium carbonate for each grain of the magnesium salt as calcium carbonate. Figure 7 shows the response of the resin t o varying concentrations of salt regenerant at three salt dosages. The operating character of the resin in waters of varying hardness is shown in Figure 8. Inasmuch as there seems t o be no definite trend towards a change in capacity with a change in degree of hardness, the data of Figure 8 are averaged and the salt efficiencies calculated for Figure 9. At a salt efficiency of 0.40 pound of salt per kilograin of calcium carbonate, Dowex 50 shows an operating capacity of 39 kilograins of calcium carbonate per cubic foot. I n softening sodium waters, Figure 10, the capacity holds up well a t a salt content as high as 120 grains of salt per gallon, except on regenerating with a low salt dosage. No essential variation in capacity of the resin was observed in softening 30-grain water when the flow was varied from 2.5 to 10 gallons per square foot per minute (Figure 11). When operating the resin in the hydrogen cycle, the course of the cycle was followed by noting the p H and degree of acidity of the effluent with a change of 10% of the titrated acidity being taken as the end point. Figure 12 presents the operation of Dowex 50 with hydrochloric acid regeneration. The definite optimum regenerant concentration here indicated is more pro-

15

G R A I N S N A C L f GAL,

Figure 10. Effect of Sodium Chloride o n Capacity of Dowex 50 in Softening 30-Grain Water

mo

INDUSTRIAL AND ENGINEERING CHEMISTRY

1354

12 L B S 15% HzS04/CU.FT. 0

p g 35

Figure 13. Effect of Sulfuric Acid as a Regenerant on Capacity of Dowex 50

25

A

mum of 650 p.p.m. The softening plant was constructed in 1928; the initial units were 9 X 16 foot horizontal tanks operated upflow nTit,h untreated greensand. In 1942 one greensand unit was replaced b y nowex 30, followed shortly in 1943 by complete conversion to Dowex 30. I n April 1946, one vcrtical unit 9 feet in diameter by 12 feet in height mas added and was filled with 300 cubic feet of D o w x 50. This unit is equipped with a Palmer surface agitator to aid in the removal during backwash of heavy accumulations of iron hydrate. This iron hydrate had been a serious problem in the horizontal units because of insufficient headroom for backu-ashing. Therefore, all units were converted into vertical softeners during 1947. In Table I11 the operating results are given for greensand, Do~vex30, and

16

N A 2 S O + P 3 0 GRAINS GAG03 P E R GALLON YGCLz 3 0 GRAINS GAG03 PER GALLON

$$

L B S OF I5%BHCL/CU.FT REGENERATION NPICL

*

-

30 P E RORA"' G A L L OcN 4c03

NAHC03-30

Vol. 40, No. 8

GRAINS GAG03

PER GALLON

NApSO4-30 G R A I N S C A C O ~ nounced than it is in the case of salt regeneration. Sulfuric acid is .PER GALLON fully as efficient a regenerant as hydroch1oric acid in the absence Figure 14. Effect Of 50 of calcium ions (Figure 13). For On Capacity Of example, pounds of 15% hydroOperating in the N a + s H + The data presented in Table 111 indicate clearly Cycle chloric acid give an operating cathe advantages gained by conversion of this softenpacity of 41 kilograins per cubic foot, and 12 pounds of l5Yc suling plant from greensand to synthetic resinous exfuric acid give 48 kilograins per cubic foot. Apparently acid changers. The cycle time has been extended 2- to 4-fold; regeneration gives a higher operating capacity than obtained in effluenthardness has been reduced: salt efficiency has been imthe sodium cycle a t the same equivalent dosage. the hydrogen-sodium cycle, ~i~~~~ 14,the operating capacproved; backwash and rinse losses are lessened; turbidity of the water is decreased; and replacement, of exchanger has been ity is related to the anion content of the sodium water. The sodium salts arranged in the order of diminishing operating capaei&minat,ed. ties are sulfate, bicarbonate, and chloride. The anomalous behavThe data sholv clearly that D ~50 is ~a definite ~ improve, ~ ior of the sodium sulfat'e x a s noted and verified in repeated tests, ment over Doffex 3o in all of these operatil~gfactors--the imbut as yet is not explained. In sulfuric acid regeneration the calcium ion tolerance is shoxyn provcment is attributable directly to the greater exchangE capacin Figure 15. In the usual downflow operating cycle calcium SUIity of Dowex 50. fate precipitation in the bed will not be removed in the acid efAt the same time these data indicate that the plant is not obfluent, causing a calcium leak and a premature end point of the taining the exchange capacity and salt efficiency from any one of run. A vigorous backmash after rinse might conceivably these three exchange materials that would be expected from labviate this difficulty. Apparently the calcium tolerance limit is about 20y0of the total cation content of the w t e r being treated. oratory tolver tests. Part of t,his discrepancy is caused by the I n the case of high calcium waters a pretreatment,of the exhausted exchanger with salt, solution would provide a gypsum-free eschange bed on regenerating with sulfuric acid.

FIELD TESTS

Doxvex 50 has been available for commercial use since December 1945. At the present time service for 22 niont,hs has been obtained on a 300-cubic foot unit inst,alled a t LIidland, hIich., for' the final softening of boiler-feed water. The influent water to this unit is Tittabamtssee River water pretreated with lime, caustic soda, and sodium aluminate, filtered through anthracite coal, and adjusted to 8.5 pH with sulfuric acid. The average hardness of the influent is 3.0 grains per gallon. The total solids content averages 330 p.p.m. with a large seasonal variation, showing a masi0

10

30

20

40

CALCIUM I N EQUIVALENT PERCENT OF TOTAL OATION CCHTEYT

Table 111.

Itern Exchanger volume Bed area Bed deoth Soft wa'ter o u t p u t Cycle time EWuent hardness Exchange capacity NaCl dosage KaCl concentration NaCl efficiency Backwash f rinse loss Turbidity in effluent, S o r m a l flow r a t e Backwash r a t e Ooeratine duration Operating duration Operating duration Replaceiuent

Figure 15. Calcium Tolerance with Sulfuric Acid Regeneration

Softening of Boiler-Feed Water at Dow Chemical Co., Midland Plant (400 pounds per square inch) GreenDomex Dowex Unit sand 30 50 Cu. f t . 300 300 300 Sq.it. 120 60 60 Inches 60 30 60 Gal. per cycle 150,000 1,000,000 2,000,000 Hours 17 8 33 P.p m. as CaCOa 5 to 7