Silica optical fibers are used more and more for delivering laser power in numerous medical applications. Many therapies require the reliable delivery of high laser power to ablate tissue. In addition, the power being delivered continues to increase in an effort to reduce procedure time and cost. Even though the majority of the fiber probes are used once and then disposed of, physicians depend on the fiber to perform without failure for the length of the procedure. In addition, the fiber should also not jeopardize patient and doctor safety. Laser damage to the optical fiber can sometimes occur when the fiber is bent while transmitting high power.

Table 1 – Properties of Fiber Tested

Kidney stone ablation is a primary example. The procedure ureteroscopy—involves passing a small medical telescope, called an ureterscope, through the urethra into the bladder, up the ureter and into a pole of the kidney to the location of the stone. A small radius bend of the fiber, sometimes down to 1cm, is often required when the stone is located in the lower poles of the kidney. A common laser used in this procedure is a Ho:YAG laser, which delivers a wavelength of 2,140nm. The typical fiber for Ho:YAG laser delivery is step-index multimode fiber that has a pure silica core with Fluorine (F)-doped silica cladding or has a Germanium (Ge)-doped silica core with pure silica cladding. The fiber has low residual hydroxyl (OH) content so the water absorption at 2.1um is low.

During this procedure, fiber fracture has been reported to occur under different laser power levels while being bent to different curvatures. Some occurred with the fiber bent to a 10mm radius of curvature, under a laser power level as low as 200mJ.[1] While others, in a 5mm radius bend diameter while transmitting kJ level laser power.

Fig. 1 – Index profile of Fibers A, B, and C.

OFS, Avon, CT, a manufacturer of optical fibers and assemblies for the medical market, examined the damage of step index multimode fibers transmitting Ho:YAG laser power, bent down to a radius to 5mm, and carrying an average power of up to 100W. The result of this investigation was the development of a manufacturing process—implemented during the draw of optical fiber—that improves the performance of an optical fiber with a silica core, silica clad, and a hard polymer coating. Specifically the coating performance was improved. The improvement is clearly seen when comparing one fiber (HCS-2) to different types of fiber—and fibers made with different manufacturing processes.

The Ho:YAG laser operates in quasi-CW (continuous wave) mode with the pulse repetition rate at 50Hz and pulse energy of 2J. The typical length of the fiber tested was approximately 5 meters. The three fibers tested were step-index multimode fibers with a core diameter of 365um, labeled Fibers A, B, and .C, as shown in Table 1.

Fibers A and B have a pure silica core and F-doped silica cladding with an NA of 0.22. Fiber C has a Ge-doped silica core and pure silica cladding with an NA of 0.29. All three fibers have UV cured fluoroacrylate coatings and ethylene tetrafluoroethylene (ETFE) buffers. The fluoroacrylate coating has a lower refractive index than that of silica. It also performs as a second cladding that guides light in the silica cladding. The index profile of the fibers is shown in Figure 1. The fluoroacrylate coating of Fiber A is labeled as HCS®-1 (trade name HCS®). The fluoroacrylate coating of Fibers B and C is labeled as HCS-2 (trade name HCXtreme®).

Table 2 – Summary of Test Results

Two types of tests were carried out to measure the fiber performance under bend and laser power. The first test measured the breaking diameter, under laser power, with a constant speed of decreased span bend. The fiber was held between two jaws of a two-point bend tester, spaced at 25mm. The laser was switched on and the plate was moved inward at a constant speed of 2mm/s. When the fiber broke, the distance between the plates was recorded as the fiber breaking diameter. Twenty to 30 measurements were done for each fiber sample. Then the breaking diameters were sorted in descending order and assigned a corresponding rank n— where n = 1, 2, 3 . . . N (N is the total number of fibers tested). The cumulative failure probability Fi is calculated using: Fi=( n - 0.5) / N. [2]

The second test measured the total transmitted power before fiber fractured when the fiber is bent to a fixed diameter: The fiber was first bent to 12mm, then the laser was switched on, and the total power (100W) was transmitted until the fiber fracture was recorded. This procedure was repeated at bend diameters ranging from 6.5mm to 5.5mm. The total transmitted power vs. bend diameter was then plotted.

Results and Discussion

Figure 2 plots the Weibull distributions of the breaking diameter with 100W of power for Fibers A, B, and C. The mechanical strength of the three fibers is very close as shown in Table 2. Because Fiber C has a larger buffer diameter and glass diameter, the median breaking diameter appeared to be larger. When under 100W of laser power, the median breaking diameter for Fiber A increased to 23.5mm and has a wider distribution. The breaking diameter of Fiber B changed 0.9mm, to 5.47mm. There is almost no change in the breaking diameters of Fiber C with and without power.

Fig. 2 – Fiber failure probability vs. bend diameter for Fibers A, B, and C with 100W average power.

The breaking diameter under power for Fibers A and B is very different even though the strength of the two fibers without power is almost identical. The only difference between the two fibers is the manufacturing process for the fluoroacrylate coating and the resulting change in properties of the coating. The indication here is that the polymer coating plays an important role in this damage mode. When optical fiber is bent, light guided inside the core of a fiber can leak into the cladding and reach the polymer layer. Because the polymer typically has a higher absorption and a lower laser damage threshold than the silica, the polymer material is more susceptible. Since the damage can initiate at the polymer layer in this failure mode, the properties of the polymer can greatly affect the performance of the fiber. For example, the index refraction of the polymer was investigated. A lower-index coating can improve the performance of single-mode fiber under bend and power.[3] Different polymer materials having similar optical properties can behave differently under high power.

HCS fiber is tested to have a higher damage threshold than other types of plastic clad silica (PCS) fiber. For 600um core polymer fiber, the HCS fiber damage threshold is >8mJ; while PCS is ~1-2mJ.[4] HCS-2 is designed and manufactured to reduce the coating absorption and stress and increase the homogeneousness thereby increasing the threshold of laser induced damage.

Fig. 3 – Bend diameter vs. total transmitted energy.

In Figure 3, we plot the total transmitted power vs. bend diameter for Fibers B and C. Fiber A is not shown in Figure 3 because it broke at 10mm instantly when t100W was launched into the fiber. The logarithm of the total energy transmitted and the bend diameter can be fitted linearly.

Fiber C can transmit more power when bent to a fixed diameter than Fiber B. When bent to 6mm, Fiber B can transmit 2kJ before breakage while Fiber C can transmit more than 20kJ without breaking. The difference between Fibers B and C lies in the fiber structures as shown in Figure 1: the NA of Fiber C is higher and the index difference of the glass cladding and the polymer coating of Fiber C is higher than that of Fiber B. Thus the bend loss from the core is lower in Fiber C than Fiber B.[5]

The result is shown in Figure 4. At a 6mm bending diameter, bend loss of Fibers B and C are 38% and 12% respectively.

Final Comments

Fig. 4 – Bend loss of Fluosil and Ultrasil fiber 272 at 2.1um with 0.22NA.

Fiber C with a lower bending loss and optimized low-index coating is better suited for laser power delivery under tight bend due to its design. The NA of Fiber C is higher, the index difference of the glass cladding and the polymer coating is also higher and, most importantly, the manufacturing process of the coating during fiber draw has been optimized. OFS has subsequently developed a full line of HCXtreme fibers, with different geometries to address a wide range of applications that require high performance silica/silica fibers to reliably deliver high laser powers.

This article was written by Jie Lie, PhD, Senior Manager of Linear Engineering & Development; Xiaoguang Sun, PhD, Fiber Development Engineer; and Adam Hokansson, Product Line Manager, at OFS Specialty Photonics Division, Avon, CT. For more information, Click Here 


  1. R.M. Percival, E.S.R. Sikora, and R. Wyatt, “Catastrophic Damage and Accelerated Ageing in Bent Fibres Caused by High Optical Powers,” Electronics Letters, vol. 36 [5]; pp414-6; 2000.
  2. FOTP-028C Measuring Dynamic Strength and Fatigue Parameters of Optical Fibers by Tension.
  3. G.S. Glaesemann, M.J. Winningham, D.A. Clark, J. Coon, S.E. DeMartino, S.L. Logunov, and C-K Chien, “Mechanical Failure of Bent Optical Fiber Subjected to High Power,” Journal of the American Ceramic Society, vol. 89 [1]; pp50-56; 2006.
  4. S.W. Allison, G.T. Gillies, D.W. Magnuson, and T.S. Pagano, “Pulsed Laser Damage to Optical Fibers,” Applied Optics, vol. 32, pp291-297; 1993.
  5. A.A.P. Boechat, D. Su, D.R. Hall, and J.D.C. Jones, “Bend Loss in Large Core Multimode Optical Fiber Beam Delivery Systems,” Applied Optics, vol. 30, pp321-327; 1991.