Believe it or not, rare earth doped optical fiber has been around for about 60 years. Eli Snitzer was the first to report on laser action in a neodymium doped silicate glass fiber emitting at 1.06 µm back in 1961 [1]. This was the first demonstration of laser action of a rare earth in a silicate glass host, as well as in the form of an optical fiber. It was a multi-mode fiber with a 300 mm core having a refractive index of 1.54 and a cladding index of 1.52 made from soft glass melts at the American Optical Company, in Southbridge, MA. The fiber was fabricated by the “Rod in Tube” method which is performed exactly as it is named, by forming a core rod and a separate cladding tube, inserting the rod in the tube and then drawing as fiber. A great deal of research was done in the next decade on material hosts for the rare earths, such as heavy metal fluoride and chalcogenide glasses as well as silica, and much work on different rare earths ions.

Interestingly enough, in this same time frame, the development of optical fiber communication systems was ongoing. Eli Snitzer, also in 1961, published a paper on the theoretical description of single-mode fiber [2]. With the right index difference and core size, a single-mode fiber could be produced; however, at that time fiber losses were on the order of 1 dB/m. In 1966, C.K. Kao and G.A Hockham published a paper that theoretically specified the requirements for a long range fiber optic communication system[3] which called for fiber optic losses to be on the order of 10 to 20 dB/km, at that time not yet achieved. In 1970, Robert Maurer, Donald Keck, and Peter Schultz at Corning, broke the 20 dB/km barrier by inventing a process to deposit ultra-pure silica on the inside of a glass substrate tube[4]. The process is called flame hydrolysis deposition where a vapor of O2 and SiCl4 is converted to SiO2 in the form of micron sized particles that deposit on a substrate. Ultra-pure glass forming precursors can be further purified using commonly known techniques based on the volatility of the precursor. The act of vaporizing the precursor and delivering it to the reaction zone is yet another purification step of the process. The achievement of low loss glass using vapor phase processes is what enabled the fiber optic telecommunications industry. Purity is the name of the game and the same holds true for rare earth doped fibers.

It was inevitable that there would be a marriage of these low loss telecommunication grade fiber fabrication processes and processes for rare earth doping of fibers. There are several variations of the flame hydrolysis process: OVD, VAD and MCVD to name a few. The rest of this narrative will concentrate on just the MCVD process, due to its flexibility, and the various rare earth doping processes that are used in combination with MCVD.

 Rare Earth Doped Fibers

 During the early 1980’s, I had the honor of working for Dr. Snitzer as part of his fiber optic team at Polaroid Corp. in Cambridge, MA. This group invented and demonstrated the first double-clad fiber laser [5,6] which led directly to the development of high-power fiber lasers that we have today. We also published some of the first reports of an erbium-doped fiber laser and amplifier [7]. In order to make these devices, we needed to have rare-earth-doped fibers and as such, needed to develop a process to do so. Delivering the rare earth to the hot zone (reaction zone) or getting it into the glass core was the issue. We developed a vapor phase process that used organo-metallics (chelates) [8] together with MCVD equipment made by SG Controls to deliver the rare earth and other precursors to the hot zone; the process had its advantages and drawbacks. A solution doping process as well as a halide vapor phase process was developed using the MCVD process at Southampton University [9][10]. To compare and contrast each of these methods requires that we look at each process in detail and describe some of the advantages and drawbacks.

Several fiber parameters are important to consider when choosing which process to use in fabricating a rare earth doped fiber. Will the fiber device be core pumped or cladding pumped? Which rare earth is being used and why? Where are the absorption and emission bands? Will OH contamination be an issue? What concentration of rare earth is needed? What NA is required for the core? Is there a particular index profile shape required? Is there a particular dopant profile required for the rare earth? What overall laser efficiency needs to be met? And finally, what production requirements need to be met and is the process repeatable? By looking at each process, we can evaluate the advantages and disadvantages with respect to each of these requirements.

The most common processes used to make rare earth doped fibers are the Rod and Tube method, mentioned in the first paragraph, the Solution Doping Process and the Vapor Phase Process. The latter two are used most commonly in conjunction with the MCVD process which will be discussed. The success of the Vapor Phase Processes is dependent on the precursors used.

Solution Doping

This technique starts by first depositing an un-sintered porous layer of silica on the inside of a silica substrate tube using MCVD. The tube is then removed from the lathe, and the porous layer is doped by filling the tube with a rare earth solution, usually a rare earth chloride dissolved in a solvent. The solution remains in the tube until it has time to fill the porous structure of the un-sintered silica layer. The solution is then drained from the tube, leaving the porous soot layer impregnated with the solution. The solvent evaporates and the rare earth chloride is left behind. The tube is then returned to the lathe where it is dried and then sintered into a glass layer. Co-dopants such as aluminum chloride and others can be introduced in the same way. The process is then repeated for a number of cycles until the desired core size and index profile is achieved.

The advantage of this process is that it is, relatively, the easiest process to carry out. It requires no modifications to the standard MCVD process and only an additional process station to fill and drain tubes with solutions and perhaps some gas drying requirements. High rare-earth concentrations can be achieved (as much as 5 wt.%) and successful fibers have been fabricated for many applications. Some of the drawbacks include: OH contamination due to the difficulty of removing hydrogen ions introduced by the solution and the difficulty in adequately drying a hydrated rare earth chloride; clustering effects that can occur due to inadequate mixing of the rare earth with the silica matrix, which depends on diffusion only; overall purity concerns due to the purity of the rare earth chlorides used, the solutions used and the constant manipulation of the tube, including removing it from the lathe and possible exposure to contaminates. In addition, it is a multi-step process that can take many cycles and several days to complete a preform. Careful control of the soot layer porosity is critical, as this can affect the dopant concentration in the glass layer. Depending on what fiber parameters are required, a solution doping process can be developed; however, it will take significant time and process development to achieve a successful fiber.

Vapor Phase Doping:

The vapor phase doping process uses a volatile rare earth precursor in conjunction with the MCVD process and delivers the rare earth to the hot zone along with the other MCVD precursors (SiCl4, GeCl4 etc.). Aluminum chloride is also delivered to the hot zone by vaporizing AlCl3. These reactants are converted to their oxides in the hot zone in the form of a soot which is then sintered into a uniform glass layer,  as is done in the standard MCVD process. The process provides for complete control of the dopant by controlling both the temperature and carrier gas flow rate, much the same as the standard MCVD process. The dopants are intimately mixed with the silica matrix as they are formed at the same time in the hot zone from the vapor. Dopant and index profiles can be controlled layer by layer and large core preforms can be made in-situ without the need of removing the tube from the lathe. Typical preforms with 2 to 3 mm cores can be fabricated in one 8-hour shift, greatly speeding up the turnaround time for the development and production of new fibers. The process is not as simple as solution doping and does require some modifications to the MCVD equipment. There are two different types of rare earth precursors that have been used, organo-metallics and rare earth halides which are discussed further.

Rare Earth Chelates

These are organo-metallic compounds which contain a rare earth ion and provide volatility for the otherwise non-volatile metal. These compounds have significant vapor pressures in the 200 °C range and can be transported to the hot zone via heated lines. They can be vaporized from the liquid or solid state. The main advantage lies with the vapor phase transport as described above, and very high concentration preforms have been fabricated using these precursors. However, the chelates have some disadvantages. The main one is the ability to obtain high-purity precursors due to the nature of the compound and how it is fabricated. Contamination from other rare earths and/or other unwanted metal ions is likely. As organo-metallic compounds, they bring hydrogen ions into the hot zone which is undesirable for maintaining low OH contamination. There are fluorinated versions of the chelates available, but these will deliver fluorine ions to the hot zone which will affect the deposition efficiency and soot viscosity and will reduce the index of the glass. These may or may not be desirable depending on the fiber requirements, but if a fluorinated compound is used, you’re stuck with it. A second disadvantage is the incompatibility with the standard halides used in the process. Standard halides are very oxidizing and at the vaporization temperatures used, there can be breakdown of the compound leading to premature deposition of unwanted material in the substrate tube prior to the hot zone. A third disadvantage is the possibility of breaking down the compound if it is exposed to too high a temperature. Therefore, temperature control along the heated lines is critical. If there is a hot spot, for example, the material will break down into a blackened clog in the delivery line. If there is a cold spot, the material will condense.

Rare Earth Halides

These are the standard halides of rare earths (RECl3). The main disadvantage to using these precursors is that high temperature (850°C to 900°C) is required to reach significant vapor pressure to deliver the rare earth to the hot zone. Engineering modifications to the MCVD process accomplishes this allowing realization of all the advantages. First, the halides are anhydrous, so no hydrogen is introduced to the hot zone making this the driest of all the processes available. For some rare earths, such as ytterbium, this may not be a big issue; however, for thulium, this is very important since Tm lases at one of the major OH absorption peaks in silica in the 1. 9 µm to 2. 0 µm wavelength range. Since the rare earth halides are compatible with the standard MCVD halides, there is no issue with premature deposition in the substrate tube. Using the halide process, it is possible to achieve very low NA (≈ 0.05) preforms with significant rare earth concentrations (≈1 wt.%) [11]. This is achieved by the ability to minimize the other dopants to only what is necessary for solubility of the rare earth in the silica matrix and not having to include index-lowering dopants that can cause high core stress. This is very important for CW high-power fiber lasers and high-peak-power pulsed fiber lasers since larger cores are desired while still maintaining single-mode operation. Very high rare earth concentrations have been achieved using the halide precursors (> 8.5 wt.%) [12]. This can be particularly important for thulium, since high concentration increases the 2 for 1 pumping of Tm at 976 nm.  And finally, a key advantage is that the halides are fabricated from the rare earth oxides which are available in five “9s” purity levels. The rare earth halide process provides the highest purity, lowest OH contamination, the highest achievable concentration of rare earth, and the best process flexibility.


Choosing which process to use in fabricating a rare earth doped fiber is a critical decision which depends on the particular requirements. Solution doping has the advantage of easiest implementation and lowest initial equipment cost with the downside of longer processing time and significant handling of the pre-collapsed preform. It can suffice assuming the core glass can meet the specifications. Vapor phase processes have higher initial costs and are more difficult to implement with the advantage of shorter processing times, no pre-collapsed tube handling and significant improvement in core glass quality. The processes must be weighed against each customer’s individual situation and requirements.

[1] E. Snitzer, “Optical Maser Action of Nd3+ in a Barium Crown Glass”, Phys. Rev. Lett. 7(12), 444-446, (1961)
[2] E. Snitzer, “Cylindrical Dielectric Waveguide Modes” Journal of the Optical Society of America, Vol. 51, No. 5, pp. 491-498, May, 1961
[3] K. C. Kao and G. A. Hockham, “Dielectric-fibre surface waveguides for optical frequencies,” in Proceedings of the Institution of Electrical Engineers, vol. 113, no. 7, pp. 1151-1158, July 1966.
[4] D.B. Keck, P.C. Schultz, Patent No. 3,711,262, Filed 1970, Issued 1973
[5] E. Snitzer, H. Po, F. Hakimi, R. Tumminelli, B. McCollum, “Double-clad, offset-core Nd fiber laser” (first report of cladding pumping), Proc. Conf. Optical Fiber Sensors, Postdeadline paper PD5 (1988)
[6] E. Snitzer, “Cylindrical Dielectric Waveguide Modes” Journal of the Optical Society of America, Vol. 51, No. 5, pp. 491-498, May, 1961
[7] H. Po, E. Snitzer, R. Tumminelli, L. Zenteno, F. Hakimi, N. M. Cho, and T. Haw, “Double Clad High Brightness Nd Fiber Laser Pumped by GaAlAs Phased Array,” in Optical Fiber Communication Conference, Vol. 5 of 1989 OSA Technical Digest Series (Optical Society of America, 1989), paper PD7.
[8] E. Snitzer, H. Po, F. Hakimi, R. Tumminelli, B. McCollum, “Erbium Fiber Laser Amplifier at 1.55 mm with pump at 1.49 mm and Yb sensitized Er oscillator” Proceedings of Optical Fiber Communication, OFC’88, New Orleans, Postdeadline paper PD2 (1988)
[9] J.E. Townsend, S.B. Poole, D.N. Payne, “Solution-doping technique for fabrication of rare-earth-doped optical fibres”, Electronics Letters, Volume 23, Issue 7, p. 329 –331 (1987)
[10] S.B. Poole, D.N. Payne, M.E. Fermann, “Fabrication of low-loss optical fibres containing rare-earth ions” Electronics Letters, Volume 21, Issue 17, p. 737 –738 (1985)
[11] V. Petit; R. Tumminelli; J. Minelly; V. Khitrov “Extremely low NA Yb doped preforms (<0.03) fabricated by MCVD” Proceedings Volume 9728, Fiber Lasers XIII: Technology, Systems, and Applications; 97282R (2016) SPIE LASE, 2016, San Francisco, California, United States
[12] R. Tumminelli; V. Petit; A. Carter; A. Hemming; N. Simakov; J. Haub, “Highly doped and highly efficient Tm doped fiber laser” Proceedings Volume 10512, Fiber Lasers XV: Technology and Systems; 105120M, SPIE LASE, 2018, San Francisco, California, United States
English English Español Español