Preparation of high purity silica

Preparation of high purity silica

Mat. Res. Bull. Vol. i0, pp. IZ63-1Z66, Printed in the United States. PREPARATION 1975. P e r g a m o n Press, Inc. OF HIGH PURITY SILICA D. A. Pi...

276KB Sizes 0 Downloads 12 Views

Mat. Res. Bull. Vol. i0, pp. IZ63-1Z66, Printed in the United States.


1975. P e r g a m o n Press, Inc.


D. A. Pinnow, F. W. Dabby, # I. Camlibel, A. W. Warner, L. G. Van Uitert Bell Laboratories Murray Hill, New Jersey 07974

(Received October i, 1975; C o m m u n i c a t e d by N. B. Hannay) ABSTRACT We describe an economical self-sustaining process for making high purity bulk silica boules by flame oxidation of silane, SiH4. We also report a simple procedure for converting the bulk silica boules into rod stock, which is a useful shape for certain applications such as preparing optical fiber preforms. The silica produced by this process has a surprisingly low water content, 59 ppm, which is over an order of magnitude less than found in commercially available silica produced by flame techniques. The high purity of our silica, which was made with house gas in an u n c o n t r o l l e d environment, can be inferred from its low measured optical absorption 5.4 dB/km, at a w a v e l e n g t h of 1.06~. Since the invention of making fiber optical waveguides with pure fused silica cores surrounded by b o r o s i l i c a t e glass claddings was made two years ago (i), several of the present authors have been involved in developing practical processes for realizing this structure. During the initial phases of this work fiber preforms were made starting with a commercially available high purity silica rod upon which a layer of borosilicate glass was deposited by flame oxidation of b o r o n and silicon hydrides (2). The composition of the b o r o s i l i c a t e glass was constant with a molar ratio of approximately 3 Si02:l B20s. Recently we developed a modified process to produce fiber preforms with a graded refractive index profile (3) since this structure is known to have a greater bandwidth capability (4). Rather than depositing a constant composition of borosilicate glass, the boron concentration was slowly increased during the course of deposition from a small initial value of 9 Si02:l B203. The fact that the low b o r o n content D. A. Pinnow is currently employed by Hughes Research #Laboratories, Malibu, California. F. W. Dabby is currently employed with Fiber Communications, Incorporated, Orange, New Jersey. 1263





Vol. I0, No. IZ

silica was well behaved led us to attempt to produce boron-free silica glass by a similar process. The m o t i v a t i o n for doing this was p r i n c i p a l l y one of economics. Commercially available high purity silica rod is quite expensive and the projected cost estimates for plasma produced silica, based on the development work of Nassau and Shiever (5), are relatively high. Making pure silica was, in fact, a straightforward extension of the previous work and the first attempt succeeded. The optical absorption of the resulting material was low and the OH concentration was much lower than we had anticipated. The method we used for making this pure silica is described b e l o w along with our evaluation of the quality of the material. Our procedure for making silica was to deposit a powdery silica layer produced by flame oxidation of silane, SiH4, onto a vitreous silica substrate. The substrate and powdery layer were subsequently heat treated to sinter the powder into high quality vitreous silica. The volume of the resulting silica piece was much greater than that of the initial substrate. If this procedure were used for production, a small fraction of the silica piece could be shaped into a new substrate to sustain the silica p r o d u c t i o n cycle, while the remainder could be used for fiber preform p r e p a r a t i o n or other applications.








The fused silica core rod serves as the substrate for deposition of powdery silica from the torch flame. SiH4 gas is injected into the flame through the central orifice of the torch.


10, N o .






Figure i is a schematic of our silica d e p o s i t i o n equipment which is located in a l a b o r a t o r y hood. A commercial rod of silica 3 mm in diameter was used as the substrate. It was held at its top and b o t t o m w i t h chucks w h i c h rotated and t r a n s l a t e d up and down by a simple m o t o r drive system. Before b e g i n n i n g the deposition a commercial glass blower's torch was used to fire polish the substrate rod as shown in the figure. This step was found to be n e c e s s a r y to assure a high quality flaw-free interface b e t w e e n the substrate rod and the deposited layer. D e p o s i t i o n was accomp l i s h e d by supplying SiH4 gas to the central orifice of the same torch that was used for fire polishing. The torch's flame was sustained by natural gas and oxygen fed through a tube bundle that surrounded the central orifice. The SiH4 was mixed and reacted with the oxygen in the flame to produce two reaction products, a fine powder of Si02 w h i c h adhered to the substrate rod and water vapor w h i c h was p r i n c i p a l l y e x h a u s t e d from the hood. Deposition continued for a p p r o x i m a t e l y 3 hours during which a p o w d e r y layer of silica a p p r o x i m a t e l y 18 mm in diameter and 6 inches long was formed on the 3 mm diameter substrate rod. By w e i g h i n g the substrate rod before and after d e p o s i t i o n and m o n i t o r i n g the SiH4 flow rate we conclude that the chemical c o n v e r s i o n e f f i c i e n c y of SiH4 to Si02 Was a p p r o x i m a t e l y 50% (6). The substrate rod with its adherent silica layer was s u b s e q u e n t l y loaded into a laboratory oven where it was heat treated in a helium atmosphere at a temperature of 1425°C for a p p r o x i m a t e l y i0 hours. During this period the powdery silica layer shrank to a diameter of i0 mm as it sintered into a boule of high quality silica glass. The heat treatment was concluded when the interface b e t w e e n the deposit and substrate was no longer visible. The deposited silica material was prepared for e x a m i n a t i o n by carefully sawing it away from the substrate rod and then shaping it using conventional procedures. One piece was ground into a rod shaped sample suitable for an optical a b s o r p t i o n m e a s u r e m e n t using the calorimetric technique d e v e l o p e d by P i n n o w and Rich (7). The a b s o r p t i o n loss was d e t e r m i n e d to be 5.4 dB/km at the 1.06~ exciting wavelength. Such low loss is indicative of high purity, although a careful analysis of the spectral v a r i a t i o n in loss should be performed before the purity level can be quantified. A n o t h e r sample was prepared into a polished plate for infrared t r a n s m i s s i o n studies. B a s e d on the magnitude of the 9.7~ absorption band we d e t e r m i n e d that the OH c o n c e n t r a t i o n was only 59 ppm (8). At :first this was a rather surprising result since commercial silica prepared by gas torch techniques has t y p i c a l l y 1900 ppm of OH. We now believe that the low OH c o n c e n t r a t i o n in our material is due to out d i f f u s i o n of OH from the p o w d e r y deposit during heat treatment. Out d i f f u s i o n appears likely because of the large surface area of the powdery deposit. In contrast, all of the c o m m e r c i a l l y prepared b u l k silica is formed at s u f f i c i e n t l y high substrate t e m p e r a t u r e s that it vitrifies as it is b e i n g deposited. There is very little o p p o r t u n i t y for OH to escape from the surface b e f o r e being covered over by more vitreous silica. A l t h o u g h the silica boule was cut and p o l i s h e d for the above studies, we have recently d e m o n s t r a t e d that such a boule can be directly drawn into rod stock u s i n g our fiber drawing machine. The procedure we used (9) was to feed the tip of the boule into





Vol. i0, No. 12

the hottest region of the annular resistance furnace mounted on the drawing machine. A thin fiber filament was pulled off of this tip and attached to a take-up drum using a procedure identical for drawing a fiber preform. The boule is then fed into the furnace at a relatively high rate and the take-up drum is rotated very slowly. As the rotation proceeds the flexible fiber filament easily conforms to the drum surface at one end while exerting a uniform downward tensile force on the boule at the other end. With this combination of high feed and slow take up the boule can be drawn down into a very uniform rod. For the present work we drew a i0 mm diameter boule down to a 3 mm diameter rod which is a useful form either for additional substrate material or for the core material for fiber preforms (2,3). In summary, we have developed a process for making high purity fused silica with relatively low OH concentration. The efficiency for converting the starting materials, SiH4 and 02, to silica is high which suggests that the process should be economical References i.

D. A. Pinnow, L. G. Van Uitert and J. C. Williams, 3,778,132 dated December ii, 1973.




F. W. Dabby, D. A. Pinnow, F. W. 0stermayer, L. G. Van Uitert, M. A. Safi and I. Camlibel, "Borosilicate Clad Fused Silica Core Fiber Optical Waveguide With Low Transmission Loss Prepared by a High-Efficiency)Process," Appl. Phys. Lett. 25, 714-715 (December 15, 1974 . m F. W. Dabby, D. A. Pinnow and I. Camlibel, "Graded Index CladFused Silica Core Fiber Optical Waveguides," presented at Topical Meeting on Optical Fiber Transmission, Williamsburg, Va., January 7-9, 1975.


F. W. 0stermayer, Jr., "On the Geometrical Optics Analysis of the Impulse Response of Graded-Index Fibers," Technical Digest of papers presented at the Topical Meeting on Optical Fiber Transmission, Williamsburg, Va., January 7-9, 1975.




K. Nassau and J. W. Shiever, results.

Bell Laboratories,


The chemical conversion efficiency has been found to increase as the ratio of the substrate diameter to torch flame diameter increases. For larger substrates we have demonstrated conversion efficiencies in excess of 70%.


D. A. Pinnow and T. C. Rich, "Development of a Calorimetric Method for Making Precision Optical Absorption Measurements," Applied Optics 12, 984-992 (May 1973).


D. M. Dodd and D. B. Fraser, "Optical Determination of OH in Fused Silica," J. Appl. Phys. 37, 3911 (September 1966).


This procedure was suggested by P. Kaiser of Bell Laboratories.