Method of Controlling the Index Profile of Optical Fiber Preforms

This invention relates to methods for controlling the refractive index profile of optical fiber preforms.

As is known, efficient transmission of optical wave energy along a multimode optical fiber is achieved by grading the refractive index of the fiber core. Thus, accurate knowledge of the index profile is necessary in order to assess a fiber's transmission properties. To this end, various techniques have been developed for measuring the index profile of optical fibers and optical fiber preforms. For example, U.S. Pat. No. 4,161,656 discloses various techniques for determining the distribution of the index modifying dopants in optical fibers and fiber preforms by observing the reaction of a fiber/preform to ultraviolet (uv) illumination. From this measurement, the index profile can be computed.

The shortcoming of this and various other measuring techniques is that they can only be employed after the fiber or preform has been made. A preferred arrangement would involve making measurements while the preform is being fabricated and, thereby, providing information with which to control the fabrication process itself. By this means, the resulting index distribution within the preform could be made to more closely approximate the desired profile. It is clear that such an approach would serve to reduce the number of unacceptable preforms produced and, thereby, reduce the cost of fabricating optical fibers.

The present invention is particularly applicable to the so-called "modified chemical vapor deposition" (MCVD) process in which a precursor, in the form of a gas containing glass forming material and suitable index modifying dopants, is caused to flow into a preform substrate tube wherein it is heated. This causes a layer of glass to be deposited on the inner surface of the tube.

The thickness of each deposited layer, and the concentration of dopants within each layer are functions of a number of parameters including the temperature of the hot zone created within the tube, the speed with which the hot zone moves along the tube, and the concentrations of the glass forming material and the index modifying dopants within the gas. If the temperature and speed of the hot zone are maintained constant, the thickness and the refractive index of each layer can be controlled by controlling the material concentrations within the gaseous precursor. Thus, in accordance with one aspect of the present invention, the material concentrations within the gaseous precursor being delivered to the preform substrate tube is monitored and controlled during the manufacture of the preform by a method characterized by the steps of: illuminating the gaseous precursor with ultraviolet radiation; measuring the intensity of radiant energy derived from said gaseous material; comparing said measurement with a reference signal and generating an error signal in response to this comparison; and changing the material concentration in said gaseous material, as required, in response to said error signal.

In one embodiment of the invention, the radiant energy measured is the uv radiation that traverses the gas. In an alternate embodiment of the invention, the dopant concentration is determined by measuring the fluorescence induced in the dopants by the uv radiation.

The control provided by the above-described technique can be considered a "coarse" adjustment. Ultimately, it is the actual thickness of each glass layer deposited within the preform substrate tube, and the dopant concentration within each layer that determines the index profile of the resulting preform. Thus, in accordance with a second aspect of the present invention, measurements are also made of the deposited layers employing an x-ray shadowgraph. "Fine" adjustments in the fabrication parameters are then made in response to these measurements. These can include modifying the flame temperature, the speed with which the flame moves along the preform, and/or changing the gas flow to the substrate tube.

In all cases, it is an advantage of the present invention that the measurements are made during the fabrication of the preform and, thus, provide a means of optimizing the index profile.

FIG. 1 shows an arrangement in accordance with a first aspect of the present invention for monitoring and controlling the index profile of an optical fiber preform;

FIG. 2 shows an arrangement for monitoring and controlling the index profile of an optical fiber preform in accordance with a second aspect to the invention;

FIG. 3, included for purposes of explanation, shows a cross section of a fiber preform and an x-ray shadowgraph produced thereby;

FIGS. 4 and 5 show, in block diagram, illustrative signal processors for use in connection with the present invention;

FIG. 6, included for purposes of explanation, shows an x-ray shadowgraph of a fiber preform having a plurality of deposited layers; and

FIG. 7 shows the application of the invention to the VAD method of fabricating preforms.

Referring to the drawings, FIG. 1 shows a first arrangement for monitoring and controlling the index profile of an otpical fiber preform fabricated by means of the MCVD process. As explained more fully in the copending application by J. B. MacChesney and P. B. O'Connor, Ser. No. 828,617, filed Aug. 29, 1979, now U.S. Pat. No. 4,217,027, and assigned to applicants' assignee, in accordance with this process a substrate tube is held and rotated (by means not shown) as a hot zone is caused to move along the tube in the direction indicated by arrow 12. Typically, the hot zone is produced by a flame 4 from a heating means 11 which is shown mounted on a track 3. Simultaneously, a gaseous precursor is introduced into tube 10 via an inlet pipe 13 which is connected to a material reservoir 14 by means of two pipes 8 and 9. The reservoir typically includes a source 15 of oxygen, a source 18 of an index modifying dopant material, such as germanium tetrachloride, and a source 19 of a host glass forming material, such as silicon tetrachloride. The oxygen source is connected by means including a first valve 6 to source 18, and by means including a second valve 7 to source 19.

These materials are typically in a liquid state and are advantageously heated by means of heaters 16 and 17. In operation, the gas from source 15 is bubbled through the index modifying dopant solution and the glass forming solution. The resulting gaseous mixtures are then fed by means of pipes 8 and 9 to a coupler 5 wherein they are combined. The combined mixture is, in turn, fed by means of inlet pipe 13 into the preform tube 10 wherein it is heated, resulting in the deposition of a layer of glass upon the inside surface of tube 10.

Of particular interest are the refractive index and thickness of each of the deposited layers. The refractive index of each layer is primarily a function of the relative concentrations of the two materials within the gaseous mixture. The layer thickness is a function of the concentration of the host glass forming material within the precursor gas, the temperature of the hot zone, and the rate of advance of the hot zone. Assuming that the temperature and speed of the hot zone are maintained constant at previously determined optimum values, the only things that need be controlled to obtain the desired profile are the concentrations of the host glass forming and index modifying materials. In the illustrative embodiment, this is accomplished by separately measuring the concentration of GeCl.sub.4 and SiCl.sub.4 in the respective pipes 8 and 9. To make these measurements, at least a section 20-21 of each pipe 8-9 is made of a material, such as fused silica, that is transparent to uv radiation, and the gas flowing therethrough is illuminated by means of a uv source 22, 23. In each case, the radiation is incident upon the gas in a direction that is transverse to the direction of gas flow through the tube section. In one embodiment of the invention, a detector 24, 25, located on the side of the section opposite the uv source, measures the intensity of the uv energy that has passed through the gas. Because the absorption of the uv energy by GeCl.sub.4 and SiCl.sub.4 varies as a function of their respective concentrations, the two intensity measurements are relevant to the thickness and the refractive index of the layer that will be deposited on the substrate tube. The detector outputs thus produced are coupled to a processor 26 which compares the measured values to stored reference values. If the difference between the measured and the stored signals indicates too low a concentration, a signal is generated which serves to open the appropriate valve to permit a greater gas flow through the appropriate material source. Alternatively, if the difference indicates too high a concentration, the error signal causes an appropriate reduction in the gas flow.

In an alternative embodiment of this aspect of the invention, the frequency of the uv light is selected to cause the particular material of interest to fluoresce. In this second embodiment, the detectors are advantageously displaced so as not to intercept any of the uv radiation. The induced visible light (vl), however, is sensed and the detected signals are coupled to processor 26 wherein an error signal is generated in the manner described hereinabove.

Thus, in accordance with this first aspect of the invention, the concentration of the index modifying dopant and the glass forming material are monitored and controlled as a means of controlling the refractive index and the thickness of the preform layers as the latter are being deposited upon the preform substrate tube.

An alternative method of controlling the preform fabrication process in accordance with the present invention involves making direct measurements of the layers as they are deposited upon the preform substrate, and varying the fabrication parameters, as required, in response to these measurements. An arrangement for making these measurements, using an x-ray shadowgraph, is illustrated in FIG. 2 which shows a portion of FIG. 1 including a substrate tube 10, heating means 11, and gas input pipe 13. Also shown are x-ray source 30 and an associated x-ray detector 31.

As explained in a paper entitled "Applicative Investigation of X-Ray Non-Destructive Inspection Tehchnique for Measurement of Core Diameters and Germanium Doping Concentration Profiles of Optical Fiber Preforms," by H. Takahashi et al, published in the Proceedings of the Optical Communication Conference, Amsterdam, Sept. 17-19, 1979, when a high-silica preform is illuminated by a uniform beam of x-rays, the intensity distribution of the emergent x-ray beam is modified in the manner illustrated in FIG. 3 which shows the preform cross section, including a single deposited layer, and the corresponding intensity distribution of the x-ray beam as sensed by x-ray detector 31. In the region external to the preform tube 10, the detected x-ray intensity is a maximum, given by I.sub.0. Over the interval between point a, located on the outside surface of tube 10 and point b, located on its internal surface, the detected intensity decreases from I.sub.0 to I.sub.1 as the x-ray path includes larger and larger chords within tube 10. Upon entering the first layer 33, the detected intensity continues to decrease but at a steeper slope due to the presence of the index modifying dopant, resulting in a discontinuity at the tube-layer interface b. Similarly, there is a sharp discontinuity in the intensity curve at the interface c of layer 33 and the gaseous precursor 34 that fills the balance of the region within tube 10.

As is evident from the intensity distribution curve thus defined, the thickness d of layer 33 is clearly defined by the distance between the discontinuities at interfaces b and c, whereas the dopant concentration is a function of the difference in the measured intensitites I.sub.1 and I.sub.2.

These measured values are coupled to a processor 32 wherein they are compared to previously stored reference values and appropriate error signals are generated. These error signals are then used in any one of a number of ways to modify the system parameters. For example, if it is determined that the layer thickness is too large, the error signal can be used to increase the rate at which the hot zone advances along the tube. If the measured layer thickness is too small, the rate of advance can be reduced. Alternatively, the error signal can be used to control the flame temperature and/or the concentration of materials in the gaseous precursor. If, for example, the layer is too thick, the error signal can be used to modify the reference signal stored in processor 26 of FIG. 1, thus modifying the control signal supplied to valve 7 in FIG. 1. This has the effect of "fine tuning" the gas flow and, in this particular illustration, serves to reduce the amount of SiCl.sub.4 in the gas mixture. Similarly, if it is found that the dopant concentration is not correct, an appropriate modification of the control signal applied to valve 6 will "fine tune" the concentration of index modifying dopant in the gas mixture.

Advantageously, the x-ray source and x-ray detector are suitably mounted on a movable stand (not shown) so as to track directly behind the advancing hot zone. By this means, the layers are constantly monitored and corrections are made continuously as each layer is deposited within tube 10. Thus, the slightest deviation in either layer thickness or refractive index is quickly detected and the appropriate corrections made.

FIG. 4 shows, in block diagram, an illustrative embodiment of processor 26 for generating the error signals that control the gas flow through valves 6 and 7. The processor comprises a pair of memory circuits 40 and 41 which store the relevant reference signals for each layer that is to be deposited. One store of signals, as in memory 41, relates to the dopant concentrations whereas the other store, as in memory 40, relates to the layer thicknesses. A signal from a counter 29, located at one end of track 3, identifies the particular layer being put down and permits the memory circuit to select the appropriate reference signal. Following each pass of the flame, the counter is advanced by the returning heater 11 and a new reference signal is provided by each of the memory circuits 40 and 41. These, along with the detected signals from detectors 25 and 24 are coupled to comparator circuits 42 and 43, respectively, in which they are compared and the error signals generated. The latter are then coupled to the appropriate controls for modifying the fabrication parameters. In the illustrative embodiment, the controlled components are flow valves 6 and 7.

FIG. 5 shows, in block diagram, an illustrative embodiment of processor 32 for generating error signals in response to signals from x-ray detector 31. The processor includes a video line scanner 50 whose output is an analogue waveshape of the type illustrated in FIG. 3. A digitizer 51 measures the amplitudes of the signals from line scanner 50 and converts the measured amplitudes to digital form for convenience. These signals are read, stored and compared in a read, store and compare circuit 52.

As indicated in connection with FIG. 3, each layer that is deposited results in a decrease in the intensity of the detected x-rays. FIG. 3 illustrated this with respect to one layer. FIG. 6, now to be considered shows a portion of the shadowgraph for a plurality of n+1 layers. As shown, each layer results in a decrease in x-ray intensity. To determine whether the dopant concentration is proper, the change in x-ray intensity is measured and compared to a reference signal that has previously been stored in the processor memory. Circuit 52 makes the required measurement by comparing the last intensity minimum with the new minimum. Thus, the minimum produced by the nth layer, and stored in circuit 52, is I.sub.n. (The latter is clearly recognized as such by virtue of the abrupt change in intensity as the scan passes into the gas filled region of the substrate tube.) As the n+1 layer is deposited, a new minimum, I.sub.n+1 is detected and compared to I.sub.n. The resulting difference is, in turn, compared in comparator 53 with a reference signal from memory circuit 54, and an error signal is generated.

The width of the n+1 layer is determined by generating a signal whose amplitude is proportional to the elapsed time, (i.e., t.sub.n+1 -t.sub.n) between the two measured minima I.sub.n+1 and I.sub.n. This difference signal is, in turn, compared in comparator 55 with a reference signal from memory circuit 56, and a second error signal is generated. As before, counter 29 identifies the layer in question.

As indicated hereinabove, these error signals can be used in a variety of ways. For example, the error signal from comparator 55, which relates to the layer thickness, can be used to control the flame temperature or the speed with which the flame advances along the preform tube. Alternatively, both error signals can be used to modify the magnitudes of the reference signals in processor 26 so as to fine tune the gas concentrations in the gaseous precursor supplied to tube 10. In any case, the effect is to provide a continuous control of the fabrication process.

An alternative means of obtaining deposition information can be obtained using uv illumination and either a uv or a fluorescence detector in place of the x-ray source and x-ray detector shown in FIG. 2. As the layers are deposited, the amount of uv radiation absorbed will increase, resulting in a decrease in the uv detector output. On the other hand, as more and more layers are deposited, the amount of fluorescence produced will increase. However, in each instance, the effect is dependent upon both the thickness of each layer and the dopant concentration within each layer. Thus, measurements made in this manner cannot be used to control the fabrication process as readily as the x-ray shadowgraph. However, it can be used to detect gross deviations in the fabrication process. For example, if a valve were to malfunction, significant changes in the anticipated level of uv or visible light (vl) would be detected. This change can be used to either alert the operator or simply shut down the equipment.

In the description of the invention given hereinabove, reference was made to the MCVD process. However, the techniques described can be readily applied to other preform fabrication processes as well. For example, FIG. 7 illustrates the application of the invention to the vapor axial deposition (VAD) method of fabricating preforms wherein the glass forming materials are deposited at the end of the preform so that the latter grows longitudinally. FIG. 7 shows a portion of a starting silica rod 60, a section of consolidated preform 61, a section of porous preform 62, a heater 63 which consolidates the porous preform as it is pulled through the heater, and two oxyhydrogen burners 64 and 65 which supply the glass forming materials. Typically, the preform is rotated as the glass is deposited to obtain an axially symmetrical distribution of material. To monitor the actual preform profile, the exposed preform end is illuminated with uv whose wavelength is selected to induce fluorescence in the index modifying dopant. If more than one dopant is used, successive measurements are made at different wavelengths or, if the wavelengths of the uv and induced vl are sufficiently different for the several dopants, concurrent measurements can be made. In either case, a source 66 of uv light is directed at the exposed end surface 69 of preform 62. The resulting fluorescence is detected by means of a video camera 67. Advantageously, the camera is focused on the surface of the preform to minimize any interference produced by the incident cloud of glass forming material from burners 64 and 65.

A line scanner 70 connected to camera 67 monitors the intensity distribution of the fluorescence and, in conjunction with a processor 71 generates error signals to control the fabrication process in the manner described hereinabove.

In the embodiment of FIG. 1, the gas mixtures were monitored separately in pipes 8 and 9. However, because the attenuation peaks for the several materials occur at different wavelengths, these measurements can, alternatively, be made after the gases have been combined in pipe 13. Similarly, inasmuch as the wavelength necessary to induce fluorescence is different for the different materials, this measurement can also be made after the gases have been combined. Thus, various different arrangements can be devised to practice the invention.