11688041 (P.T.A.B. Oct. 4, 2012)

Ex Parte Koyata et al

Patent Trial and Appeal BoardOct 4, 2012

11688041 (P.T.A.B. Oct. 4, 2012)

UNITED STATES PATENT AND TRADEMARKOFFICE UNITED STATES DEPARTMENT OF COMMERCE United States Patent and Trademark Office Address: COMMISSIONER FOR PATENTS P.O. Box 1450 Alexandria, Virginia 22313-1450 www.uspto.gov APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO. 11/688,041 03/19/2007 Sakae Koyata P35804 9309 7590 10/04/2012 GREENBLUM & BERNSTEIN, P.L.C. 1950 Roland Clarke Place Reston, VA 20191 EXAMINER DEO, DUY VU NGUYEN ART UNIT PAPER NUMBER 1713 MAIL DATE DELIVERY MODE 10/04/2012 PAPER Please find below and/or attached an Office communication concerning this application or proceeding. The time period for reply, if any, is set in the attached communication. PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE ____________________ PATENT TRIAL AND APPEAL BOARD ____________________ Ex parte SAKAE KOYATA, YUICHI KAKIZONO, TOMOHIRO HASHII and KATSUHIKO MURAYAMA ____________________ Appeal 2011-009285 Application 11/688,041 Technology Center 1700 ____________________ Before: FRED E. McKELVEY, ROMULO H. DELMENDO and MARK NAGUMO, Administrative Patent Judges. McKELVEY, Administrative Patent Judge. DECISION ON APPEAL Appeal 2011-009285 Application 11/688,041 2 Statement of the case Sumco Corporation (“applicant”), the real party in interest (Brief, page 3), 1 seeks review under 35 U.S.C. § 134(a) of a non-final rejection dated 12 July 2010. 2 The claims on appeal have been twice rejected. 3 The application was filed in the USPTO on 19 March 2007. 4 The application on appeal claims priority of (1) application 11/345,009, filed 5 31 January 2006 (now U.S. Patent 7,288,207) and (2) Japanese patent application, 6 filed 31 January 2005. 7 The application has been published as U.S. Patent Application Publication 8 2007/0184658 A1. 9 In support of prior art rejections, the Examiner relies on the following 10 evidence. 11 Ohta et al. “Ohta 1” U.S. Patent Application Publication 2006/0240748 A1 26 Oct. 2006 filed 10 Feb. 2004 Ohta et al. Ohta 2 Japanese Patent Application JP 2003-036872 24 Sept. 2004 Ishimura et al. “Ishimura” Abstract of Japanese Patent Application JP 08-287370 28 Apr. 1998 Appeal 2011-009285 Application 11/688,041 3 Ohta 2, a Japanese-language document, is prior art under § 102(b). The 1 Examiner relied on Ohta 1 as being an accurate translation into English of Ohta 2. 2 Answer, page 3. Applicant does not challenge the accuracy of the translation. 3 Ishimura is prior art under § 102(b). 4 We additionally cite the following evidence. 5 Nakano et al. “Nakano” U.S. Patent 6,110,839 29 Aug. 2000 Shibasaki et al. “Shibasaki” U.S. Patent Application Publication 2003/0152506 A1 14 Aug. 2003 Alexander et al. “Alexander” The Solubility of Amorphous Silica in Water, 58 Phys. Chem. 453-455 1954 The invention 6 1. Field of the Invention 7 The invention relates to an etching liquid for controlling a silicon wafer 8 surface shape. Specification, page 1, first ¶ 0003. 9 2. Description of the Related Art 10 According to applicant, manufacture of a semiconductor silicon wafer 11 includes an etching step. A wafer can have a damaged layer, i.e., a “work-12 affected” layer on the surface of the wafer. The work-affected layer is said to 13 Appeal 2011-009285 Application 11/688,041 4 induce crystal defects which in turn are said to decrease mechanical strength of the 1 wafer, further causing adverse effects on the electrical characteristics of the wafer. 2 Accordingly the defects should be removed. Specification, second ¶ 0003. 3 An “etching” is performed to remove the work-affected layer. In the etching 4 process, either an acid etching method or an alkali etching method can be used. In 5 either etching process, a plurality of wafers is dipped into an etching bath 6 containing an etching liquid where the work-affected layer is said to be chemically 7 removed. Specification, ¶ 0004. 8 The invention on appeal relates to an alkali etching bath. 9 Alkali etching is said to have advantages. Alkali etching liquid can include 10 potassium hydroxide (KOH) and sodium hydroxide (NaOH). According to 11 applicant, alkali etching has the advantages because alkali etching is said to 12 progress on the basis of diffusion controlled conditions. Despite these advantages, 13 in alkali etching there can occur “facets” whose partial depth is several μm, and 14 whose size can be several ten μm. The facets are said to deteriorate the wafer 15 surface roughness—which has been a problem in the art. Specification, ¶ 0006. 16 3. Embodiments of the invention 17 Applicant’s etching liquid comprises silica powder “dispersed” uniformly in 18 an alkali aqueous solution. The etching liquid is said to affect metal impurities. 19 Specification, ¶ 0031. 20 Applicant’s etching liquid is obtained by (1) mixing silica powder to an 21 alkali aqueous solution adjusted to a specified concentration at a specified rate, 22 (2) stirring the mixture, and (3) dispersing the silica powder uniformly in the alkali 23 aqueous solution. The alkali aqueous solution may be potassium hydroxide (KOH) 24 Appeal 2011-009285 Application 11/688,041 5 and sodium hydroxide (NaOH). Preferred is a 40 to 50 weight % sodium 1 hydroxide aqueous solution. Further, it is said to be preferable to add the silica 2 powder to the 40 to 50 weight % sodium hydroxide aqueous solution at a 3 concentration of 1 to 100 g/L. According to applicant, after the etching process, 4 the high wafer flatness degree can be easily maintained and wafer surface 5 roughness is reduced. 6 The preferred average particle diameter of the silica is preferably in the 7 range of 50 to 5000 nm. If the average particle diameter is below 50 nm, there is 8 said to occur a nonconformity with respect to metal impurities in the alkali aqueous 9 solution. Hence, we believe applicant is discouraging the use of a diameter below 10 50 nm (0.050 μm). On the other hand, if the average particle diameter is over 11 5000 mm (5 μm), nonconformity is said to occur because silica powder dispersed 12 in the alkali aqueous solution becomes massed together. Specification, ¶ 0032. 13 The more preferred average particle diameter of the silica powder is said to be 14 500 nm (0.5 μm) to 3000 nm (3 μm). Specification, ¶ 0032. 15 The overall wafer-making process steps are shown in Fig. 1. The etching 16 portion of the wafer-making process is described in ¶ 0039 and Fig. 5 (reproduced 17 below). 18 Appeal 2011-009285 Application 11/688,041 6 1 Fig. 5 depicts an etching process 2 In step 1 (shown on the left), wafers 41a are dipped into etching liquid 42a in 3 etching bath 42 where the “work-affected” layer is said to be removed. Thereafter, 4 in step 2 (shown in the middle) the etched wafers are dipped in a rinse liquid 43a in 5 rinse bath 43. Step 3 (shown on the right) is a drying step. 6 Claims on appeal 7 Claims 1-3 and 6 are on appeal and, as set out in the Claims Appendix of the 8 Brief, read as follows: 9 Claim 1 10 An etching liquid for controlling a silicon wafer surface shape, 11 wherein silica powder is dispersed uniformly in a 40 to 50 weight % sodium 12 hydroxide aqueous solution. 13 Appeal 2011-009285 Application 11/688,041 7 Claim 2 1 The etching liquid of claim 1, comprising the silica powder in an 2 amount of l to 100 g/L [grams per liter] of the sodium hydroxide aqueous 3 solution. 4 Claim 3 5 The etching liquid of claim 1, wherein the average particle diameter of 6 the silica powder is 50 to 5000 nm. 7 Claim 6 8 The etching liquid of claim 2 wherein the average particle diameter of 9 the silica powder is from 500 nm (0.5 μm) to 3000 nm (3 μm). 10 Examiner’s rejection 11 In the Answer, the Examiner has maintained a rejection of claim 1-3 and 6 12 (all the claims on appeal) as being unpatentable under § 103 over Ohta 2 and 13 Ishimura. We refer to Ohta 1 which is in English. 14 Analysis 15 Examiner’s rejection 16 The Examiner relies on ¶ 0020 of Ohta 1, which along with ¶ 0019, read: 17 According to the invention, first of all, an acid fumed 18 silica dispersion is prepared in the first step. Incidentally, it is 19 desirable that the relative surface area of the fumed silica to be 20 used is 50 to 200 m2/g. 21 Then, in the second step, an aqueous basic substance 22 solution is prepared. The aqueous basic substance solution is 23 controlled in concentration and volume by mixing with the 24 Appeal 2011-009285 Application 11/688,041 8 [acid] fumed silica dispersion prepared in the first step so that 1 the intended abrasive composition has a pH of 8 to 12 and a 2 silica concentration of 10 to 30 wt %. The aqueous basic 3 substance solution contains at least any of ammonium 4 hydroxide, sodium hydroxide, potassium hydroxide, calcium 5 hydroxide, barium hydroxide or magnesium hydroxide. 6 According to the Examiner, the matter described in ¶ 0020 “would read on 7 claimed silica powder [which] is dispersed uniformly in an alkali aqueous 8 solution.” Answer, page 3. 9 The Examiner recognizes that Ohta 2 does not describe 40-50 weight % 10 sodium hydroxide. Brief, page 9; Answer, page 3. Applicant goes on to say that 11 Ohta 2 describes a pH of 8-12 whereas the pH of the etching liquid of Claim 1 12 would be 14—or even higher. Brief, page 11. We do not find it necessary to 13 address applicant’s speculative “even higher” argument. However, a 40-50 14 weight % limitation appears in Claim 1 and it is not apparent where that limitation 15 is described in Ohta 2. We do not find the Examiner’s analysis based on Ishimura 16 to be sufficient in resolving the difference between Claim 1 and Ohta. Since an 17 element of Claim 1 is missing from the prior art, the evidence does not support a 18 rejection under § 103. Claims 2, 3 and 6 are narrower than Claim 1 and therefore 19 Ohta 2 likewise does not describe an element of those claims. 20 The Examiner’s § 103 rejection is reversed. 21 New rejection 22 Applicant previously called the Office’s attention to a published Nakano 23 Japanese patent application 08-218503 (publication 09-129624 dated 16 May 24 Appeal 2011-009285 Application 11/688,041 9 1997), accompanied by a machine translation. Nakano, cited above, appears to be 1 a U.S. patent corresponding to the Nagano published Japanese application. We 2 refer to U.S. Nakano as opposed to the machine translation of the Japanese 3 Nagano. The U.S. Nakano patent is prior art under § 102(b). 4 Claims 1-3 and 6 are rejected as being unpatentable under § 103(a) over 5 Nakano, Shibasaki, and Alexander. 6 Unlike Ohta 2, which deals with an abrasive composition (¶ 0013-0020), 7 Nakano describes etching liquids used in an etching step in a process of making a 8 semiconductor wafer. Nakano seeks to purify the etching liquid to remove, or at 9 least minimize, metallic impurity ions in an alkali solution used as an etching 10 liquid. Col. 1:7-11. 11 After what they characterize as extended research, the Nakano inventors 12 found that metallic impurity ions in an alkaline solution can be easily neutralized. 13 Col. 2:9-11. 14 Furthermore, according to Nakano, semiconductor wafers etched with the 15 alkaline solution in which metallic impurity ions were neutralized or removed 16 apparently are not as prone to deterioration. Col 2:11-15. 17 Nakano’s object was to provide a method of purifying an alkaline solution 18 which enables metallic impurities, especially metallic ions in the alkaline solution 19 to be neutralized at a low-cost and efficiently, as well as a new method of etching 20 semi-conductor wafers without deteriorating wafer quality. Col. 2:17-23. 21 The Nakano process comprises the steps of “dissolving” metallic silicon 22 and/or silicon compounds in an alkaline solution and neutralizing metallic ions in 23 the alkaline solution. Col. 2:24-32. 24 Appeal 2011-009285 Application 11/688,041 10 Suitable metallic silicons include polysilicon and single crystal silicon. 1 Suitable silicon compounds include silica and silicates. According to Nakano, the 2 preferred silicon compounds are those having high purity. Col. 2:33-37 3 The amount of metallic silicon “dissolved” in solution is not significant as 4 long as the effect of the Nakano invention is achieved. A suitable amount is 5 0.2 g/liter or more. When the amount “dissolved” is too small, it is said that the 6 resulting effects of the solution on the wafer are not enough. When the amount 7 “dissolved” is too large, it is said that it becomes economically disadvantageous. 8 Col. 2:39-46. 9 Moreover, Nakano advises those skilled in the art that the amount of the 10 silicon compounds “dissolved” is likewise not critical as long as the effect of this 11 invention is achieved. The amount of silica is suitably 5 g/liter or more. When the 12 amount “dissolved” is too small, the resulting effects of the solution on the wafer is 13 said to not be enough. When the amount “dissolved” is too large, it is said that it 14 becomes economically disadvantageous. 15 Nakano Example 4 (col. 5) is relevant to the facts in this case. Example 4 16 reads in part (col. 5:39-47): 17 To a sodium hydroxide solution (45%, 2 liters and 25 ºC.), 1 wt % 18 of silica was added. Before and after adding the silica, 10 ml of the 19 sodium hydroxide solution diluted to 45 times was sampled, 20 respectively. Then a nickel ion concentration and an iron ion 21 concentration thereof were analyzed by an ion-exchange 22 chromatography. The results of the analyses are shown in FIG. 4[, 23 reproduced below.] 24 Appeal 2011-009285 Application 11/688,041 11 1 Fig. 4 2 Fig. 4 depicts a bar graph showing the relation between 3 an iron ion concentration and a nickel ion concentration in 4 the sodium hydroxide solution before and 5 after dissolving silica as discussed in Example 4 6 7 It should be noted that the Y-axis is logarithmic. 8 Nakano goes on to say (col. 5:47-50): 9 As is apparent from the results of FIG. 4 [reproduced above], both the 10 iron ion concentration and the nickel ion concentration were 11 remarkably decreased by adding the silica. 12 While Nakano uses the term “dissolved” and applicant uses the term 13 “dispersed”, it is readily apparent that Nakano and applicant are using the same 14 ingredients (45% NaOH vis-à-vis 40-50% NaOH and silica) to prepare an etching 15 solution having the same wafer etching utility. Moreover, applicant describes and 16 claims (Claim 2) a concentration of NaOH (1 to 100 g/L) which overlaps with that 17 described by Nakano (“suitable 5 g/L or more”—see col. 2:51). At the silica 18 concentrations described as suitable by Nakano and those of Claim 2 on appeal, 19 silica does not appear to be “dissolved”. See, e.g., Alexander, page 454, col. 1 20 Appeal 2011-009285 Application 11/688,041 12 (“Method of Determining Soluble Silica”) and in particular the first full paragraph 1 in col. 2 indicating that solubility is relatively independent of the amount of solid 2 silica in a suspension, except when less than 0.1% silica powder is present in a 3 “dispersion.” It does not appear that either applicant or Nakano call for the use of 4 less than 0.1% silica. 5 On the current record before the PTO, it appears that applicant and Nakano 6 prepare similar, if not identical, etching liquids. Under the circumstances, it is 7 incumbent on applicant to establish that a difference exists, and if a difference 8 exists, that any difference would have been unpredictable. In re Best, 562 F.2d 9 1252, 1255 (CCPA 1977) (“[w]here, as here, the claimed and prior art products are 10 identical or substantially identical, or are produced by identical or substantially 11 identical processes, the PTO can require an applicant to prove that the prior art 12 products do not necessarily or inherently possess the characteristics of his claimed 13 product”); In re Spada, 911 F.2d 705, 708-9 (Fed. Cir. 1990) (where claimed 14 composition and prior art composition appear to be the same, PTO may require 15 applicant to prove there is a difference); In re Klosak, 455 F.2d 1077, 1080 (CCPA 16 1972) (inventor must show that the results the inventor says the inventor gets with 17 the invention are actually obtained with the invention and it is not enough to show 18 results are obtained which differ from those obtained in the prior art—any 19 difference must be shown to be an unexpected difference). 20 Claim 2 calls for a concentration of 1-100 g/L. Nakano describes a 21 concentration of “suitable 5 g/liter or more”. Col. 2:49-51. Applicant claims a 22 concentration falling squarely within the concentration described by Nakano. 23 Appeal 2011-009285 Application 11/688,041 13 Claims 3 and 6 call for silica having certain particle sizes. One skilled in the 1 art attempting to use the Nakano invention would have used a particle size which 2 accomplishes Nakano’s purpose, one of which is to minimize impurities. 3 Applicant too is concerned about metal impurities in the alkali solution. 4 Specification, ¶ 0032 (on page 11, six lines from the bottom). Applicant outlines 5 problems which might arise if a particle size smaller or larger than the claimed 6 particles sizes is used. One skilled in the art would practice the Nakano invention 7 in a manner to achieve Nakano’s result. It follows that one skilled in the art would 8 have been inclined to use a particle size which would obtain that result. In 9 determining the particle size to be used in the Nakano environment, one skilled in 10 the art no doubt would have uncovered the same problems found by applicant and 11 would have been able to determine an appropriate particle size to avoid those 12 problems. Moreover, notwithstanding the discussion in the Specification, 13 applicant is in a poor position to argue that particle size is “critical” given that 14 Claim 1 contains no particle size limitation. Lastly, we note that silica particle 15 sizes within the scope of applicant’s size are known. Shibasaki, col. 1:63 16 (0.1-1 μm, a range of particle sizes which overlaps applicant’s 0.5-3 μm). 17 Decision 18 Upon consideration of the appeal, and for the reasons given herein, it is 19 ORDERED that the decision of the Examiner rejecting claims 1-3 and 20 6 over Ohta and Ishimura is reversed. 21 FURTHER ORDERED that, on the current record before the Board, 22 Claims 1-3 and 6 are unpatentable under § 103 over Nakano, Shibasaki, and 23 Alexander—a new rejection under 37 CFR § 41.50(b). 24 Appeal 2011-009285 Application 11/688,041 14 FURTHER ORDERED that our decision is not a final agency action. 1 FURTHER ORDERED that within two (2) months from the date of 2 our decision, appellant may further prosecute the application on appeal by 3 exercising on of the two following options: 4 Option 1: Request that prosecution be reopened by submitting 5 an amendment or evidence or both. 37 CFR § 41.50(b)(1). 6 Option 2: Request rehearing on the record presently before the 7 Board. 37 CFR § 41.50(b)(2). 8 FURTHER ORDERED that no time period for taking any subsequent 9 action in connection with this appeal may be extended under 37 CFR 10 § 1.136(a)(1)(iv). 11 REVERSED 12 (New rejection under 37 CFR 41.50(b)) 13 bar 14 .June, 1054 SOLUBILITY OF AMORPHOUS SILICA IN WATER 453 0.08 0.06 THE SOLUBILITY OF AMORPHOUS SILICA IN WATER - d i / / - BY G. B. ALEXANDER, W. M. HESTON AND R. K. ILER Grasselli Chemicals Department, Experimental Statim, E . I . du Pont de Nemours and Company, Inc., Wilmington, Delaware Received November 9, 1966 The solubility of amorphous silica in water a t 25’ is shown to involve an equilibrium between the solid phase and a mono- The same solubility (0.01 to 0.012%) was found using a suspension The increase in total dissolved silica above The concentration of Si(OH), meric form of silica in solution, presumably Si(OH),. of finely divided amorphous silica powder and sols of colloidal particles of silica. p H 8 is shown to be due to the presence of [ H B S ~ O ~ ] ion in addition to Si(OH)r in solution. in equilibrium with the solid phase apparently is not affected by p H . Introduction.-Numerous studies of the solu- bility of silica in water have been reported1-’ (Fig. 1) but it was not clear in most cases whether solubility- equilibrium was actually established. Also it was not determined whether the silica in apparent solu- tion was present as monosilicic acid, polysilicic acid or colloidal silica particles. It is the purpose of this investigation to clarify these points. In order to ensure that equilibrium was reached, we have approached the true value of the solubility a t 25’ from the unsaturated state by starting with suspensions of colloidal particles of amorphous sil- ica in water, and from the supersaturated condition by starting with soluble silicic acid of low molecular weight, letting it polymerize to the colloidal state. In order to determine whether the silica in solution is monomeric or polymeric, we have employed a modification of previously described methods in- volving the formation of silicomolybic acid.*-’ By this method it is possible to measure the concen- tration of monosilicic acid in the presence of other forms of silica, since the monomer. reacts with mo- lybdic acid very rapidly; polysilicic acids react more slowly and colloidal forms of silica scarcely at all. Monosilicic acid is known to result from the dissolution of silica and in solution has been shown to be a hydrated form of Si02,13 presumably Si(0H)d. The solubility has already been shown to increase above pH 9.14 It was another objective of our investigation to explain this behavior. It is well known tha.t amorphous silica is more soluble than crystalline silica (quartz) but it has not been demonstrated that the different forms of amor- phous silica such as finely divided powder, wet silica gel and colloidal particles in the form of a sol, are solubIe to the same extent. Accordingly, these different forms of silica were employed in this study. The fact that we find the solubilities of these dif- (1) FIAT Rev. Ger. Sci . , Part 1, 265 (1939-1946 (Pub. 1948)). (2) C. Struckman, Liebigs A n n . Chem., 94, 341 (1855). (3) C. S. Hitchen, Inst. of Mining and Met . , 255 (1935). (4) V. Lenher and H. B. Merrill. J . A m . Chem. Soc., 39, 2630 (1917). (5) L. U. Gardner, Anier. Inst. of Metallurgical and Mining Engi- (6) R. Spychalski, A . annorg. allyem. Chcm., 239, 317 (1938). (7) C. M. Jephcott and J . H. Johnston, Arch. I n d . H u g . and O c i cupational M s d . , 1, 323 (1950). (8) F. Dienert and F. Wandenbulcke, Compt. rend., 176, 146 (1923). (9) R. W. Harmon. THIS JOURNAL, 31, 616 (1927). (10) T. Okwia, J . Chem. SOC. Japan , P w e Chem. Sect., 72, 927 (1951): C. A . , 46, 6995 (1952). (11) E. Weitz, H. Franck and M. Schuchard, Chem. Ztg., 1 4 , 256 (1950). (12) G. Jander and K. F. Jahr, KoEloid-Beih., 41, 48-57 (1934). (13) H. and W. Brintzinger. 2. anorg. ollgem. Chem., 196, 44 (1931). (14) 0. Jander and W. Heukeshoven, ibid.. 201, 301 (1931). neers, Technical Publication No. 929, p. 7. I I I I 0 100 200 Temp., O C . Fig. 1.-Solubility of amorphous silica: 8, Struckman*; 9, Gardner6. 0, Spychalskie; 0, Hitchens; 0, Lenher and Merril14; 0, jephcott and Johnson.’ ferent forms to be the same, indicates that the fun- damental structure of amorphous silica in these various forms is identical. This agrees with the conclusions reached by other investigators,15-IS who employed X-ray and electron diffraction methods. Moreylg concluded that if there are any “crystals” in amorphous silica glass, they must be scarcely larger than one unit cell of cristobalite. The same appears to be true for silica sol and gel particles prepared in aqueous solutions. Morey points out that since a crystal is, by definition, a regular repetition of unit cells, it is artificial to call materials (15) 0. E. Rsdoze!vski and H. Richter, Kolloid Z., 96, 1 (1941). (16) L. Krejci and E. Ott, THIS JOURNAL, 35, 2061 (1931). (17) J. T. Randall, H. P. Rooksby and B. 9. Cooper, J . SOC. Glass (18) B. E. Warren, J . A p p l . Phzls., 8, 645 (1937); Z. Krist . , 86, (19) G. W. Morey, “The Properties of Glass,” A.C.S. Monograph Technol., 15, 54 (1931). 349 (1933). Series No. 77, Reinhold Publishing Corp.. New York. N. Y., 1938. 454 G. B. ALEXANDER, W. M. HESTON AND R. K. ILER Vol. 58 “crystalline” in which crystals are only one unit cell in size. Experimental The following experimental work was designed to deter- mine the solubility of amor hous silica in the forms of (A) fine powder, (B) colloidal sof)ution, and (C) freshly prepared silica gel, and to elucidate the nature of the soluble silica in acidic, neutral and basic aqueous solutions. It will be noted that the silica structures used in this study had very small particles (high surface area); this factor has been shown to be important if equilibrium conditions are to be achieved in a reasonable period of time. 0.5 0.1 4 0.05 “0 0.02 8 3 cs 0.01 m 0.005 0.001 2 4 6 8 10 12 PH. Fig. 2.-Solubility of silica in water: 0, sols by removal of Na+ from sodium silicate solution; 0 , suspensions of “Silica A,” pH adjusted with HC1 or NaOH; 0, calcu- lated from equation 1. Method of Determining Soluble Silica.-It has been verified recentlyz0 that the observation of Weitz, Franck and Schuchard” is valid: monomeric silicic acid reacts very quickly with molybdic acid, while polysilicic acids react more slowly. Under the reaction conditions which were adopted for this study, at least 98% of the monosilicic acid in a solution reacts with molybdic acid within two minutes.” The molybdate reagent is prepared as follows: 100 g. of (NH~)eMo,Ozt.4HzO(J: T. Baker)isdjssolvedindistilled water and is diluted to one liter. Just, prior to use, 40 ml. of this reagent is added to 860 ml. of distilled water and 100 ml. of 1 N sulfuric acid. This strength molybdate reagent is mixed with enough sample to result in a maximum soluble silica concentration of about one milligram per 50 ml. in the color developing solution. A Beckman Model DU photoelectric quartz spectrophotometer with a thermostati- cally controlled cell (25’) is used to follow the increase in optical density of the sample due to the formation of the yellow silicomolybdic acid complex. The optical density is measured at a wave length of 400 mp with a slit width of 0.04 p . Under these conditions, one milligram of silica per 50 ml. of solution produces an optical density of 0.720. As soon as the silica sample is mixed with the molybdate re- agent, the color reading is noted a t l/Z-minute intervals until a constant value is obtained. (A) Solubility of Silica Powder.-As a typical silica powder, an amorphous silica, “Silica A,” produced by com- bustion of silane vapors, was employed. This material contained 99.9% silicon dioxide and had a specific surface area of 240 square meters per gram. This material was suspended in boiled, freshly distilled water which had been previously adjusted to pH 5.6 with a trace of HCl, and kept in waxed bottles. Concentrations of the powder in suspen- sion ranged from 0.03 to 1 .O% by weight. Different concen- trations of the suspensions from 0.03 to 1% Si02 were made (20) G. B. Alexander, J . Am. Chsm. SOC., TO, 5655 (1953). to check the effect of the amount of silica in suspension on the equilibrium solubility. The concentration of soluble silica In the supernatant liquid in these suspensions was determined after 20, 39 and 62 days, and found to increase and become essentially constant after 20 days providing at lea& 0.3% SiOz was present in the suspension. The equilibrium solubility proved to be 0.014% Si02 in solution as monomeric silica. The solubility of this silica was relatively independent of the amount of solid silica in the suspension, except when less than 0.1% silica powder is present In the dispersion. In another set of experiments, the solubility of “Silica A” in water was measured a t various pH values from 1 to 10.2. The pH was adjusted with HCl or NaOH, and 1% by weight of the silica powder was suspended in the water. The amount of silica in solution was determined after three weeks and six months at 25’. The final solubilities a t dif- ferent pH values are included in Fig. 2. The color reaction of the soluble silic: with molybdic acid was complete within two minutes a t 25 , indicating that the soluble silica was monomeric. Another series of sus ensions of “Silica A” was aged for six months in more alkayine solutions. The results shown in Table I indicate that above about pH 10.6, the solution con- tains not only monosilicic acid but low molecular weight polymerized silica, presumably present as polysilicate an- ions. TABLE I SOLUBLE S~LICA AND pH OF 1% SUSPENSION OF “SILICA A” IN NaOH SOLUTION AFTER 6 MONTHS AT 25’ Molar A8 As “active” ratio monomer, polymer,” SiOz/N6zO PH % 5% 33 10.28 0.053 0.00 16 10.60 . l l . 00 8 10.85 .21 .05 3 11.04 .36 .36 a Low molecular weight polysilicate ions, which react with molybdic acid reagent within 10 minutes. Rate of Reaction of Molybdate Reagent with “Silica A.” -Two milliliters of a 1% suspension of “Silica A” aged six months in methanol was added to 100 ml. of the molybdic acid solution and the development of the color was followed a t 25”. Time, min. 1 40 Apparent % Si02 in methanol soln. 0.0007 0.0014 This not only shows that very little silica is dissolved in the methanol, but also that the “Silica A” is essentially not reactive with the molybdate reagent even after 40 minutes. (B) “Soluble Silica” in Aged Sols.-Instead of approach- ing solubility equilibrium by suspending silica powder in water or dilute alkali, it is possible to start with a colloidal solution of silica particles or polysilicic acid and determine the concentration of monomeric silica in such colloidal solu- tions. (1) Alkaline Sols.-Clear sols of colloidal silica were pre- pared by partial removal of sodium ions from a dilute solu- tion of sodium silicate by means of an ion-exchange resin. At the outset, these sols were supersaturated with respect to monosilicic acid. Equilibriuy was then approached by aging the sols for 6 months at 25 . In this preparation a commercial sodium silicate (du Pont No. 9 Grade) containing a weight ratio of SiOz:NazO of 3.25: 1.0, was diluted with distilled water to produce a solu- tion containing 1.5 g. Si02 per 100 ml. Acid-regenerated “Nalcite” HCR resin, washed and then air-dried for twp days, was added in weighed amount to a sample of the di- luted silicate solution. The mixture was stirred for 3.0 minutes at 25O, and then filtered to remove resin. The amount of resin required to remove the desired amount of sodium in each experiment was predetermined in prelimi- nary experiments. The final molar ratio of SiOn: NazO in the silica sol was then determined by analysis. The diameter of the particles of colloidal silica after aging for six months is not known, but it was probably no greater than 5 or 10 mp, since the sols were very clear. To deter- . June, 1954 SOLUBILITY OF AMORPHOUS SILICA IN WATER 456 mine the concentration of monosilicic acid in these sols, aliquots of said sols were reacted with molybdic acid re- agent,. The rate and extent of reaction were used to identify the concentration of monosilicic acid. Because the size of the colloidal silica particles in the sols was very small, t>hey did not interfere in the colorimetric rcaction. I t will be noted in Table 11, that for SiO3:NazO rat.ios above 3, the “active” polymer concentration is higher than in Table I, indicating that true equilibrium solubility was not quite reached. However, a t higher ratios (77 and 224) the con- centration of monomer corresponds essentially to t’he solu- bility of aniorphous silica. TABLE I1 % SILICA IN SOLUTIOS IN 1 .5% SILICA SOL AFTER 6 MONTHS AT 25” AS DETERMINED BY THE SILICOMOLYBDATE ME HOD Molar --pH- AB ratio After 6 AB “active” SiO?: Nap0 Initial months monomer polymer 3 .25 10.95 11.03 0.32 0.34 25 8.73 9 . 2 ,017 .03 77 7 .30 8 . 6 .012 .03 224 6.45 7 . 5 ,012 .02 I t is significant that the 3.25 ratio solution shown in Table I1 contained about the same amount of sodium polysilicate (“active” polymer) as the 3 ratio sol in Table I. Also it will be noted that the amount of silica in true solution below pH 9 is comparable to that of “Silica A.” A plot of “soluble” silica, as indicated by the silica which reacts with the molybdate reagent wit,hin two minutes, versus pH, is shown in Fig. 2 for “Silica A” and for the sols of colloidal silica. 2) Acid Sols.-A solution of silicic acid containing 1% si62 , prepared by ion-exchange from 3.25 ratio (SiOz: NazO) sodium silicate and having a pH of 2.7 adjusted with HzSOI, was aged for 1 month at-25”. A t the end of this time, poly- merization had progressed to the point where the concen- tration of molybdate-reactive SiOz had dropped tp 0.03%. Samples of this clear aged sol were then diluted with HzSO~ to give solutions containing 0.1% Si02 and ranging in pH from 2.1 to 4.0; the concentration of molybdate reactive silica was followed over a period of 12 days. The concen- tration of monosilicic acid in the diluted sols increased from an initial value of 0.003 to 0.010% in every case. While the solubility had probably not reached equilibrium after 12 days, it ap eared to be approaching the range of 0.01 to 0.015 found for “Silica A” in aqueous solution. ( C ) Solubility of Silica Gel.-In a concurrent study of the reaction of low molecular weight silicic acid with molyb- dic acid,20 a silica gel was repared by polymerizing a solu- tion of silicic acid (I .O% &OZ) a t pH 2.7 and a room tem- This gel was then de- f: ydrated under vacuum at about 90’ and slurried in dis- tilled water. The solubility of silica was determined by the silicomolybdate method. A concentration of 0.02% Si02 was found in the supernatant solution. That this soluble silica was monomeric was shown by the fact that it reacted completely with molybdic acid within two minutes. Fur- ther evidence that the soluble silica was of low molecular weight was obtained by measuring the freezing point of the solution using a technique described previously by one of the authors.21 The freezing point depression of this silica solution was 0.012’ lower than the freezing point of a solu- tion free from silica but containing a concentration of HZ- so4 (0.001 N ) equivalent to that in the silicic acid solution. The deprpsion for 0.02% SiOz as monomer is calculated to be 0.005 ; agreement is within the accuracy of the tem- perature measurement. Discussion Establishment of Equilibrium.-From the fore- going results, it is clear that solubility equilibrium at 26” has been established. This is shown by the fact that when amorphous silica powder (“Silica A”) was suspended in water the concentration of monomeric silica increased to a value of 0.014% Silica in s o h . after 6 months, % erature of about 25” for two weeks. (21) G. B. Alexander, J . Am. Chem. Soc., 7S, 2887 (1953). SiOz, while in a solution of polysilicic acid super- saturated with monosilicic acid which was permitted to polymerize for six months (Table 11), the con- centrat’ion of monomer decreased until it was about the same value, i e . , 0.01270 Si02. The solubility of silica in these systems was relatively constant at about 0.012-0.014% in the pH range 5 to 8 (Fig. 2). At DH 2.1 to 2.7. the solubilitv of monomeric silica frdm polymerized polysilicic ”kid is approxi- mately 0.010% SiOz. The Effect of pH on the Solubility of Silica.-The increase in the total “soluble” silica a t high pH can be explained on the basis of the following Guiiib- rium, assuming that the concentration of Si(OH)r does not change with pH. Si(OH)r + (OH-) = (H0)sSiO- + HzO Although silicon in the silicate ion is represented in the above equation to have a coordination number of four, it is believed that it may actually have a coordination number of six in accordance with the formula Si(OH)e-z, which would be analogous to the fluosilicate ion. The equilibrium constant for the above equa- tion can be calculated from data obtained by Roller and Ervinz2 in a study of the association of silicate ions in the CaO/SiOz/H20 system. These authors found a t 30” - 10-8.8 [H+ [(HO)aSiO-] [Si(OH)41 Let St = total solubility of silica including monomeric silicic S, = concn. of monosilicic acid in g. Si02 per 100 ml. Whence acid and silicate ions, in g. Si03 per 100 ml. [H+l[St - S m l - [S,1 Taking S m = 0.012%, then the value of St for various pH values was calculated. As shown in Fig. 2, the values which were obtained experimen- tally are in reasonably good agreement with those calculated. This is evidence that the concentra- tion of Si(0H)o in equilibrium with the solid phase is not affected by pH. While it is certain that the solubility of silica in- creases a t high pH because of the formation of sili- cate ion in addition to &(OH), in solution, it is also possible that a t low pH the solubility may increase by reaction of Si(0H)d with acids, especially when a trace of fluoride ion is present, to form silicofluoride anions. The possibility that acids other than hydrofluoric may promote the solution of silica is indicated by Jander and Henke~hoven,’~ who found that in acid solution the solubility of silica from silica gel was highest in 0.1 to 0.01 normal solutions of HC1 or “ 0 3 , where a solubility of 1.5 X molar solu- tion (0.0009%) of silica was found. A combina- tion of chloride ion with silicic acid is indicated by Sadek,23 who measured pC1 with a silver chloride electrode and found that chloride ion is bound to the silicic acid in the proportion of one molecule of HCl per molecule of SiO2. (22) P. 6 . Roller, Jr., and Guy Ervin, Jr., ibid., 62, 468 (1940). (23) H. Sadek, J . Ind. Chem. Soc., 29, 507 (1952).