AU1477783A

AU1477783A - Improved manufacture of optical fibers

Info

Publication number: AU1477783A
Application number: AU14777/83A
Authority: AU
Inventors: Arthur David Pearson; Kenneth Lee Walker
Original assignee: Western Electric Co Inc
Current assignee: AT&T Corp
Priority date: 1982-04-12
Filing date: 1983-03-16
Publication date: 1983-11-04
Also published as: GB2118165A; EP0105327A1; CA1188102A; JPS59500512A; EP0105327A4; WO1983003600A1

Description

REDUCINGTHETAPERINANOPTICALFIBERPREFORM

Background of the Invention

1. Field of the Invention This invention relates to an improved method of making optical fibers, by reducing the taper in optical fiber preforms.

2. Description of the Prior Art

In the production of optical fibers, a preform i initially made, from which an optical fiber is subsequentl drawn. Some techniques for making a preform include the step of depositing silica, possibly including dopants, ont the inside surface of a tube. One commercially successful "echnique is described in U. S. Patent No. 4,217,027, coassigned with the present invention, which describes the modified chemical vapor deposition (MCVD) technique. In the MCVD technique, glass precursor material, typically including SiCl^ and an oxidizing medium, are flowed throug a rotating tube. A moving external heat source heats the rotating tube while traversing the length of the tube. The resulting moving hot zone in the tube causes oxidation of the glass-forming precursors, and deposition of reacted material upon the inside of the tube downstream from the moving heat source. Then, the moving heat source heats and consolidates the previously deposited material to produce a glass layer. However, because the silica material tends to deposit downstream from the heat source, there tends to be less material deposited where the reactants enter the tube (referred to as the entrance of the tube) . The material slowly increases in thickness at downstream portions of the tubes. This is referred to as the "taper" of the deposit, and as the effect is more pronounced near the entrance of the tube, the term "entrance taper" is also frequently used in the art. The tube and deposited material is then collapsed, and an optical fiber drawn from the preform. An optical fiber typically comprises a core surrounded by a cladding, with one or both being deposited inside the tube by the MCVD or other process. A taper will produce an axially nonuniform cross-sectional area of the 5 core after a fiber is drawn from the preform. This nonuniformity can, in some cases, produce a degradation in the transmission qualities of the fiber. Therefore, the entrance taper portion of the preform, which typically comprises about 10 to 20 percent of the length of the

10 preform, is not used in making an optical fiber. Clearly, this is not_. a desirable situation, and steps have been proposed for reducing the entry taper. One proposal for reducing the taper has been to modify the flow of reactants into the tube as the torch moves along the length of the

15 tube. However, this typically is difficult to control and can result in unintended shifts in the composition of the deposited material, particularly when dopants, for example germanium, phosphorus, boron, fluorine, etc., are included in the gaseous reactants. Other techniques have kept the

20 flow rate of gases constant but have varied the velocity of the torch as it traverses the tube. Slowing the velocity of the torch will result in relatively more material being deposited in the vicinity of the torch. Various "ramping" profiles, by which the torch velocity is slowed down at the

2.5 entrance of the tube, have been proposed. Linear ramping can reduce the taper to some extent, but the entry taper still remains larger than is desirable. Step functions and exponential functions have been proposed. Improved methods of reducing the taper, including the entry taper, in an

30 optical fiber preform are desirable. Summary of the Invention

This invention is a method of reducing the taper in an optical fiber preform by providing for an initial heat source velocity that is relatively low near the

35 entrance of the tube, which then increases as the heat source proceeds along the tube, and then decreases for a portion of the traverse of the tube. The velocity then either remains constant, or increases as the heat source continues toward the end of the tube. The proper velocity profile can be determined by a heat zone deposition function, which relates the rate at which a material is deposited along a length of the tube in the vicinity of a torch or other heat source in relation to the position of the heat source. Once the heat zone deposition function is at least approximately determined, it is used to calculate a heat source velocity that reduces the taper. The deposition thickness is then experimentally determined using this heat source velocity function. If the taper is still excessive, an error function is calculated which is then used to improve the heat source velocity function. By one or more successive approximations, a heat source velocity profile is obtained which minimizes taper, including entrance taper, in an optical fiber preform. Brief Description of the Drawings

FIG. 1 shows a typical preform deposition technique; FIG. 2 shows a typical heat zone deposition function for a torch;

FIG. 3 and 4 show torch velocity, and resulting deposition depth as a function of position along a tube, using a constant velocity and also using a profile of the present technique;

FIG. 5 shows a typical heat zone deposition function for a plasma-assisted deposition process. Detailed Description

The following detailed description refers to a method of making an optical fiber preform wherein the torch (or other heat source traversing the length of a preform tube) velocity is varied to achieve a more constant deposited layer thickness. Referring to FIG. 1, a stream of reactant gases enters a tube, typically made of glass, that serves as the substrate for the deposition of optical fiber material. This tube is frequently referred to as the "support tube" in the art. Normally, both cladding layers and core layers are deposited on the inside of the tube. However, in some techniques, the tube itself serves as a cladding, with the deposited material becoming substantially only the core of a resulting optical fiber. The difference between the cladding and the core is normally one of dopants included in the reactant gases used for producing the inner core that raise the index of refraction of the deposited glass. Alternately, or additionally, dopants can be included in the cladding portion that lower the index of refraction of the deposited glass, as compared to pure silica.

Referring to FIG. 1, precursor gases (10) are flowed into a rotating tube (12). A torch (11) is used for reacting the gases, causing subsequent deposition on the inside of the tube. The position of the center of the flame of the torch is located a distance Z_t from the entrance of the tube.. The highest deposition rate for the precursor materials will be at distance Z_maχ along the tube downstream from the torch. The distribution of deposited material that forms due to the torch-induced reaction is referred to herein as the "torch deposition function" and is illustrated qualitatively in FIG. 2. Notice that the distribution of material falls off more slowly to the right (downstream) side of Z__πaχ, due to the flow of reactant gases in the direction shown in FIG. 1. While the deposition is shown being entirely downstream of Z_fc (i.e., where Z-Z_fc>0) in FIG. 2, deposition can also occur upstream. This is because Z_t is defined as the distance to a fixed point in the heat zone, but a broad heat zone (for example) can deposit some material upstream of the defined point.

In the present invention, the torch deposition function is first determined, being expressed as a function of the form:

d(Z-Z_fc) (1) where d is the local deposition rate at position Z when the torch is at Z_t. (All positions are measured from the entrance end of the tube, where the reactant gases are introduced into the tube.) The deposition rate can conveniently be expressed in units of grams per minute per unit of tube length. When the torch deposition function is determined, it is inserted in Equation (2):

D(Z) = /* d(Z-Z_t) ^-y Z. (2)

D(Z), the total deposition function, gives the total amount of deposit at axial position Z, where V is the torch velocity as a function of torch position.

The torch deposition function can be determined in a variety of ways. One method is to simply flow precursor gases through the tube and hold the torch at a given position for a period of time. Then, the amount of deposited material in the vicinity of the torch is measured and the deposition function determined directly. That method suffers from the disadvantage that because the torch is held stationary, the temperature profile on the tube will be determined solely by the stationary torch. In practice, in producing an MCVD preform, the torch is moving along the tube so that some portions of the tube have been heated more recently than others due to the torch traversing the length of the tube and returning. Therefore, the actual tube wall temperature profile will not be as simple as that indicated by a stationary torch. The presently preferred method for determining the torch deposition function is to first deposit a quantity of material in the tube by traversing the length of the tube a number of times with the torch, typically at a constant velocity. To measure the thickness of deposited material along the tube, it is convenient to immerse the preform into an index-matching fluid, illuminate the preform from behind, and scan across the diameter of the preform with a microscope. The microscope reveals the precise point of interface between the deposited material and the fluid inside the preform, and hence the thickness of the deposit. A scan at least every 2 cm in the entrance region, and about every 5 cm along the rest of the length of the preform, has been found to yield adequate accuracy. For the index-matching fluid, a glycerine-water mixture is satisfactory. This measurement gives D(Z), the total deposition function, for the velocity profile used. This measured value of D(Z), which includes an entrance taper, can then be substituted in Equation (2). Then, d(Z-Z_t) is solved for in Equation (2), with V(Z_fc) being known. Then, d (Z-Z_fc) is used to optimize the velocity profile for reduced taper. This is accomplished by setting the total deposition function D equal to a constant, referred to as D and using the function d(Z-Z_t) determined above; see Equation (3) .

D* = * d(Z-Z_t) 1 Z. (3)

Then, the velocity V is solved for, in Equation (3) , with the resulting velocity being referred to as V . In the next step, a preform is then made using the velocity V . Then, the resulting thickness of the deposited material is measured (as before) and is referred to as D^E. An error function is defined, referred to as D ; see Equation (4).

** * E DCZ) = D -D Z)* = - (Z-Z_t) V Z^) dZ. (4)

** *

Note that D «D over most of the length of the tube. From Equation (4) is calculated a new velocity correction factor, referred to as V¹¹. A preform is then fabricated using a new corrected velocity, as given in Equation (5).

This process is repeated as many times as is necessary to achieve a velocity function that produces the uniformity of deposition depth to within a desired value. Usually, the error function decreases rapidly, and only 1 to 3 iterations are typically required.

The above-noted technique was used to reduce the taper in an optical fiber preform, as further discussed in the following Example.

EXAMPLE An optical fiber preform was made using a support tube having an inside diameter of 19 mm, and outside diameter of 25 mm, and a length of about 90 cm. Flowing through this tube was SiCl^ at 12 grams/minute, O2 at 2 liters/minute. He at 6 liters/minute, and in addition, small amounts of SiF^ and POCI3 were added for achieving a desired refractive index. The preform was made using a total of 15 passes of the torch to deposit material on the inside of the tube at a deposition rate of

2.3 grams/minute. Initially, a constant torch velocity was used, as shown at 41 in FIG. 4. This produced a tapered deposit radius, as shown at 31 in FIG. 3. Next, the iteration procedure discussed above was used to optimize the torch velocity. Two iterations were performed, as indicated by Equations (2)-(5) above, with corresponding measurements being made of the deposition thickness for each iteration in order to construct the next correction factor for the velocity. The resulting torch velocity is shown as line 42 in FIG. 4, which produced a resulting profile shown as. line 32 in FIG. 3. Tt can be seen that a substantial reduction in the taper, and especially a reduction in that portion near the entrance of the tube, has been achieved; the entrance taper length was only about 10 cm. Further iterations of the above-noted technique can be used to further reduce the taper. Referring to FIG. 4, there are several features of significance in the velocity profile 42. Near the entrance of the tube, the torch initially starts out at a relatively low velocity in portion 43. In some cases, this initial velocity may be approximately zero. In a second region 44 of the curve, the velocity initially increases over that of region 43, and then decreases. This is a significant departure from prior art velocity profiles, which typically called for an increase in the torch velocity as a function of distance. Next, the torch enters region 45 wherein the velocity again increases. In some cases, the velocity increase may be small; that is, approximately zero in region 45. If only an optimized linearly increasing torch velocity had been used, it is estimated that the entry taper in the above Example would be about 30 cm.

The above process has been described mainly in terms of a single torch which produces a single heat zone. However, other preform fabrication processes are also possible. For example, a plasma process using a tandem heat zone is described in ϋ. S. patent application Serial No. 143,834, filed April 25, 1980, and coassigned with the present invention. In that technique, species of which may or may not be an MCVD process, a tandem heat zone is provided with zone I being produced by a plasma and zone II typically, although not necessarily, produced by a torch for consolidation of particles that are produced. For a tandem heat zone, a deposition function such as shown in FIG. 5 can result, with other shapes being possible.- In that Figure, Z_maχ is the position of maximum deposition due to the torch, and Z_maχ2 is that due to the plasma fire ball. In addition, water cooling of a heat zone can be provided, such as shown in U. S. Patent No. 4,302,230. The use of water cooling can affect the torch deposition function. However, a torch deposition function can be determined whatever the nature of the heat source applied. Note that the term "torch" is used illustrative in this context, and the heat source may be by other means. Thus, the term "heat zone distribution function" generally describes the distribution of particulate matter due to reactions by precursor materials in a heat zone, whether produced by a torch or otherwise.

While the heat zone distribution function has been determined above by using an iterative technique based upon the deposition depth of a trial preform, and with the further possibility of determining the deposition function directly by the use of a stationary torch, another possibility is to determine the deposition rate based upon the temperature of the wall of the tube and the temperature at which the reaction occurs. A thermophoretic model can then be used to estimate the deposition of particles along the tube; see, for example, "Thermophoretic Deposition of Small Particles in the Modified Chemical Vapor Deposition (MCVD) Process," by K. L. Walker et al, in the Journal of the American Ceramic Society, Vol. 63, pages 552-558 (1980). See also "Thermophoretic Deposition of Small Particles in Laminar Tube Flow," by

K. L. Walker et al, in the Journal of Colloid Interface Science, Vol. 69, pages 138-147 (1979). Other methods of modeling a deposition process, and thereby calculating the heat zone deposition function, may be used as desired by persons of skill in the art.

It is furthermore apparent that based upon the above observation, that a decrease in heat zone velocity, following a rapid increase in velocity, is typically required to obtain a reduction in taper, it is apparent that an empirical technique can be utilized to directly determine the appropriate torch velocity without the use of the heat zone deposition function. However, the use of the heat zone deposition function has been found to significantly aid in determining the appropriate heat zone velocity according to the iterative process noted above.

^V

Claims

1. A method of making an optical fiber by steps comprising directing precursor material through a tube and reacting said precursor material in a heat zone and depositing reacted material on the inside of siad tube, with a heat source traversing substantially the length of said tube at a velocity that is a function of the position of said source along said tube, and thereafter collapsing said tube to form a solid rod, and drawing an optical fiber from the rod,

CHARACTERIZED IN THAT the velocity of said heat source comprises a first velocity region in the vicinity in the entrance end of said tube into which said reactants are introduced, and a second velocity region downstream of said first velocity region that is initially at a higher velocity than any velocity attained in said first velocity region, and with the velocity thereafter decreasing in said second velocity region, and a third velocity region downstream of said second velocity region, wherein the velocity is nondecreasing, whereby the length of the entrance taper in said tube, defined as the region in which the thickness of the deposited material in the vicinity of the entrance of the tube is substantially less than the thickness of the material in downstream portions of the tube, is substantially reduced as compared to the entrance taper in a tube made at a constant torch velocity.

2. A method of making an optical fiber by steps comprising introducing reactant material into the entrance of a tube, reacting said material by means of a moving heat source, and depositing reacted material on the inside of said tube, with said heat source traversing substantially the length of said tube, collapsing said tube to form a solid rod, and drawing an optical fiber from the rod, CHARACTERIZED IN THAT

-the velocity of said moving heat source axially along the tube is a function of the position of said heat _*

- 11 -

source along said tube so that the equation

is substantially constant over a maximum length of said tube, where D(Z) is the total deposition thickness of said reacted material at a distance Z along said tube, d(Z-Z_fc) is the heat zone depostion function, and V(Z_t) is the velocity of said heat source along said tube.

3. The method of claim 2 wherein the heat zone deposition function

_ d(Z-Z_t)

is determined by steps comprising depositing an amount of material on the inside of a tube at a given velocity and thereafter determining the amount of material so deposited as a function of the position along said tube.

4. The method of claim 3 whereby an error function is defined as the difference between a calculated value of D(Z) and a value determined by measuring the deposition thickness at points axially along the tube, and thereafter calculating a new velocity function V(Z_t) that reduces the differences in deposition thickness axially along the tube.

5. An optical fiber made according to the method of claim 1, 2, 3, or 4.