If your company manufactures optical preforms – or is considering doing so – then you know the gas delivery system is critically important, because it controls the precise delivery of high-purity gases and chemical vapors. Prior to joining Fiber Optic Center, much of my career at 3M was focused on operating MCVD systems. As you know, MCVD systems are extremely complicated, and every aspect must be tightly controlled. I’ve spent 24 years maintaining these systems, so I understand the many issues that can arise and how to correct them.
One way to ensure high yields and reproducibility is to initially select the gas delivery system that best meets your needs. If you currently have an MCVD system, there are ways to improve your gas delivery system or implement modifications to enhance your high yields and reproducibility.
Comparing the two types of gas delivery systems: stainless steel and Teflon/glass
As with any technology, there are pros and cons when comparing different types of systems. When it comes to gas delivery systems, the pros (or cons) can really add up, because you cannot risk having variability in preforms. Variability represents unusable fiber with significant loss. In the following paragraphs, I’ll describe specific points in the gas delivery process, and how each system – stainless steel and Teflon/glass – positively or negatively impacts that point in the process.
The first part of the gas delivery system (essentially the back of the system) is where the carrier gases (such as oxygen) and other reactive gases enter. With both types of systems, all valves and tubing are stainless steel to prevent leaks and minimize OH contamination. Typically, you do not see corrosion issues in this area for either system, because a potentially wet carrier gas will not yet be introduced to the reactive chemicals.
Moving to the next part of the system, there’s another set of valves, which may be stainless steel or Teflon block valves. Here, you’re introducing the carrier gases to reactive chemicals. Stainless steel valves work very well to prevent leaks, and can easily be helium leak-tested. However, significant problems can arise when the incoming gases have issues. For example, if your purchased gas has high water content or a gas dryer fails, the H2O and chemicals form reactive HCl, corroding the stainless steel. In many cases, you may not discover the corrosion until the final fiber is drawn and measured with high losses. It is very time-consuming to track down corrosion in the system. You may need to replace several sections of stainless steel tubing in the reactive area as well as stainless steel bubblers and chemicals. Your repair project can become quite extensive. Having stainless steel at this point in the delivery system requires a higher discipline. Purchasing dry gases and raw materials is critical. Operators must take care when installing reactive gas cylinders. Properly cross-purging regulators and valves to prevent introducing moisture is critical. A mistake in this area can lead to additional SS tubing corrosion. It takes additional discipline to ensure the stainless steel system delivers high-purity vapors to your deposition tube.
With a Teflon/glass gas delivery system, a wet gas stream will also introduce OH to the deposition tube, causing high water peaks in the finished fiber. The Teflon/glass will not be contaminated and will only require a dry-down period for recovery. Typically, repairing the problem – correcting a gas dryer or replacing a problematic chemical – will get you back up and running.
The bubbling process: Pros and cons of the 2 gas delivery systems
Let’s continue to follow the path of the gases. You may use a reactive gas directly, such as chlorine or boron trichloride, or may bubble a dry oxygen stream through chemical reagents such as SiCl4, GeCl4, or POCl3 (commonly used chemicals).
The bubbling process is critical, and it is important that the gas flow and liquid reagent temperature remain constant. In most systems, the chemical level changes in the bubbler over time, which enhances the cooling effect and reduces bubble path length. This changes the vapor pressure, reducing vapor concentration delivered to the deposition tube. It is very helpful to be able to observe the bubbler chemicals. In a stainless steel gas delivery system, you cannot observe the bubbler chemical color or level. A color change may indicate delivery system corrosion. If the automatic level control malfunctions, the incorrect liquid level may not be detected. On the other hand, in a glass bubbler, you can view chemicals, confirm proper level and confirm a uniform bubble stream. Over the years, I have found chemical visibility to be a real advantage.
Glass bubblers offer these additional benefits:
- The bubble stream can be observed and frit plugging can be detected, rather than assuming all is well.
- POCl3 is not compatible with stainless steel for long durations. If you use this liquid and have a stainless steel gas delivery system, you may want to install a glass bubbler to create a hybrid system.
- On rare occasions, you may want to increase the chemical temperature in the bubblers above 40 degrees C. This practice can pull metal ions from stainless steel and into your bubble stream. Clearly, this can degrade the fiber’s optical properties and should be considered when choosing an MCVD system.
On a related note, when the chemical level changes, the bubble path length is reduced and increased reagent cooling is seen. Both mechanisms result in a reduced vapor pickup. Companies like SG Controls Ltd. have incorporated an isothermal heater into a bubbler dip tube. If the carrier flow rate is increased quickly (for example, 100 CC to 2,000 CC), a temperature drop of the bubbler chemical is seen. The immersion heater quickly restabilizes the temperature in less than 4 minutes (typically one clearing pass) to prevent variation in vapor concentration. This small device can have a significant and positive impact on your final fiber. Depending on only the recirculating bubbler oil bath will require long restabilization times and, most likely, will result in variability in deposition rate.
As you know, bubblers must be refilled to maintain the same vaporization rate. This is another example of the pros and cons of the two types of gas delivery systems. How do you refill the bubbler without introducing moisture? If you have a gas delivery system, you know the bulk refill system is a separate cabinet – a remote dry box (with the Teflon/glass system) – that stores large containers of liquid chemicals. To refill the bubbler, you initially purge the stainless steel or Teflon delivery line with dry N2 to vent to ensure dryness. If a nitrogen gas dryer was to fail, you may accidentally introduce moisture into the refill system. As previously discussed, stainless steel delivery lines and the bubbler may be contaminated due to corrosion. In this scenario, the Teflon/glass system would again be preferred over stainless steel.
Additional pros, cons, and considerations …
- Cross-mixing gases – As gas delivery valves are opened and closed within mixing blocks, cross-mixing of vapors is possible if valves are not properly maintained. While this situation is uncommon, it does occur. It is a possibility with the Teflon/glass system and less likely to occur with the stainless steel system.
- Helium leak-test – If you manufacture optical preforms, you are likely familiar with helium leak-testing stainless steel systems. It is also possible to helium leak-test Teflon systems. The general assumption is that, with Teflon being more porous, you may detect some helium without a leak present. In reality, it takes time for helium molecules to diffuse through Teflon. You do need a regular leak-test procedure with Teflon/glass systems; although uncommon, there is a slightly higher probability for a leak. Manufacturers of Teflon/glass systems, like SG Controls, put their Teflon piping and valves in a dry box. A gas dryer delivers very dry nitrogen to the dry box to maintain a very dry positive pressure atmosphere. This prevents any moisture from diffusing through the Teflon into the vapor delivery lines over time. If you have a Teflon/glass gas delivery system (or are looking to purchase one) you should have a dry box.
- Changing chemicals – If you have a stainless steel gas delivery system, it is important to change the chemicals more frequently. Instituting a regular and frequent maintenance schedule will help you quickly discover any corrosion or discoloration. With the Teflon/glass system, corrosion is less likely, and chemical changes are seldom needed.
No matter which gas delivery system you use, there’s always room for improvement
I’d like to briefly address two additional points, since they are vitally important to the process and product quality. If you do not have these in place and would like assistance, please call.
- Modify your system with a high-quality rotary seal – It takes a lot of work to generate gas streams that are uniform and reproducible, with no variation. But if your rotary seal leaks you may lose part the gas stream. If the wetted parts of your rotary seal are stainless steel and leak, you may introduce moisture, which can cause corrosion and result in high loss problems. In Teflon/glass systems, if the rotary seal leaks and introduces moisture, corrosion will be prevented, but you may see a higher OH peak in measured fibers. Common rotary seal designs, where the shaft slips into a housing and seals on the inside diameter of the rubber O-ring, typically wear quickly. This, of course, impacts the quality of the seal and may force frequent rebuilds to prevent contamination. A better solution is available. SG Controls’ rotary seal O-rings seal with the O-ring side against an adjustable glass pressure plate. In this design, there is extremely low wear on the O-ring. This is a fantastic rotary seal that works with both types of gas delivery systems. Over the years, I have seen no corrosion issues and minimal leakage issues with this type of rotary seal.
- Calibration is critical; implement automated procedures to keep your finger on the pulse at every step – Let’s say you think your flow rate is 200 CC per minute, but you have not calibrated the mass flow controller. You manufacture a large quantity of preforms, and your mass flow controller fails. You buy and install a new controller and set the flow for what you thought it was. However, it turns out your actual flow rate was not 200 CC per minute. The preform cannot be reproduced. You need to start over, experiment, and change flows until you achieve the desired optical properties. Meanwhile, time and materials are lost. Knowing your absolute flow rate – and all other measurements – at all times is critical. Do everything you can to keep your finger on the pulse at every stage of the process. Over the years, I have designed and implemented a wide variety of manual and automated testing/calibration procedures to gain the highest level of assurance possible at every step in the process.
Conclusion
I’d like to leave you with this thought: Imagine you are standing at the corner of a busy city street. Do you wait for the crosswalk signal to turn green, so you can safely cross the street and reach your destination? Or do you plunge into the intersection while the crosswalk signal is red, despite the uncertainties and hazards? In my 25 years of operating and maintaining MCVD equipment to manufacture optical preforms, I’ve learned that it’s best to take advantage of all safety precautions and time-tested strategies.
If you are looking to select and purchase an MCVD system that will meet your needs while offering high yields and reproducibility, I’m happy to offer guidance. If you already manufacture optical preforms, call if you’d like to discuss modifying your gas delivery system or implementing calibration procedures to enhance your high yields and reproducibility.
Larry’s definition of frit:
FRIT – In the Teflon/glass gas delivery system used to manufacture optical preforms, the glass frit restricts the flow of oxygen to create a uniform bubble stream.
