1. Background

Archambault et al. [1] first proposed, in 1993, the Type I and II classification scheme, as a convenient means to describe two different classes of fiber Bragg grating (FBG). Specifically, this classification scheme was used to distinguish between FBGs written with either a low or high refractive index contrast, Type I referring to the former, Type II to the latter. The Type I gratings were written using relatively low pulse energies from a pulsed Excimer laser. As the pulse energy was increased the authors observed a sharp transition to a new regime, termed Type II, where the index contrast increased by two orders of magnitude. The Type II gratings were reported to be highly scattering, exhibited significant coupling to cladding modes and were relatively temperature insensitive.

A grating sub-class, Type IIA, was introduced by Niay et al. in 1994 [2]. While conventional Type I gratings were observed to red-shift during inscription, Type IIA gratings would be either wavelength invariant or would blue-shift slightly during fabrication. Type IIA gratings were also reported to be relatively temperature insensitive compared to Type I gratings, implying that the glass lattice is modified in a different manner.

The Type I and II classification has recently been adopted by a new community of researchers to classify waveguides written inside both glasses and crystalline materials using femtosecond laser irradiation. However, the additional requirement that conventional waveguides consist of a core which has a higher index than the surrounding cladding has led to attempts to redefine this classification in order to match the various manifestations of femtosecond laser inscribed structures. For example, Type I waveguides have been defined in many reports specifically as a positive index change [3] typically induced by exposure to low fluence femtosecond laser radiation. However, it is important to note that Archambault et al. did not assign a relative sign for the refractive index change to either Type I or II gratings, the classification was based on the relative amplitude of the index modulation only. Further, Type II waveguides are typically defined as those induced by a stress field in between two lines of damage. It should be noted that the original demonstration of these waveguides [4] explicitly acknowledged the Type II classification consistent with Archambault et al. A definition based on the structure of the waveguide does not adequately capture other forms of femtosecond laser written waveguides. For example, exposure to low pulse energies can produce either a positive or negative index change depending on the focusing conditions and the type of glass irradiated. A negative index change can still be used to fabricate waveguides when configured as a depressed cladding [5]. Similarly, optical damage features associated with exposure to high pulse energies can also be configured into a depressed cladding arrangement to produce a waveguide. Depressed cladding waveguides of these types were recently proposed as a 3rd class of waveguide [3, 6] and given a new classification, namely Type III waveguides [7].

The Type I and II classifications originally proposed by Archambault et al. [1] were based on different forms of material modification associated with exposure to either low or high pulse energies. The subsequent Type IIA classification is also linked to a specific form of material modification, one that produces unusual growth dynamics during FBG inscription. By comparison, the adaption of the Type I and II classifications for femtosecond laser written waveguides, and the introduction of the new Type III classification, represents a departure from the origins of this classification scheme based on material modification to one based on waveguides and their geometric form. This has created some confusion within the FBG and waveguide writing communities. Another shortcoming of this waveguide geometry based adaption of the Type I and II classification is that Anti-Resonant Optical Waveguides (ARROWs) [8], which exploit a cladding from positive index contrast modifications to form a photonic bandgap, fall outside of the new definitions. Moreover, due to the 3D capability of the femtosecond direct-write technique, one could imagine creating other elaborate guiding structures based on fiber geometries such as leakage channel fibers [9] or chirally-coupled-core (CCC) fibers [10]. Hence, the large geometrical variety of guided wave structures as shown in Fig. 1 would make a geometry based classification complicated and cumbersome.

 

Fig. 1 End-on microscope images of various waveguide geometries. The inscription laser was incident from the top in all images. The scale bars correspond to 10 µm. (a) Waveguide from a type I modification in boro-aluminosilicate glass exhibiting regions of positive as well as negative index change. (b) Stress-based double-line waveguide from type II modifications in titanium doped sapphire. (c) Depressed cladding waveguide in ZBLAN glass based on smooth type I modifications. (d) White light guiding in a microstructured waveguide in ZBLAN using type I modifications of negative index contrast. (d) ARROW waveguide under white light illumination in boro-aluminosilicate glass based on positive index contrast type I modifications.

Download Full Size | PDF

Clearly, the variety of designs, material modification scenarios or its final states, accessible via direct femtosecond modification of transparent dielectrics, is far richer than the simple two fold division introduced in [1], a factor that has led to attempts to introduce new categories.

2. Conclusion

We propose a return to the original intent of the Type I and II classifications, one intrinsically linked to material modification associated with working below or above the optical damage threshold of the irradiated substrate (glass or crystal), optical damage in this case referring to the regime where the target substrate undergoes a rapid change in refractive index in response to a monotonically varying incident laser pulse energy. In addition, it should be acknowledged that Type I modification can include both positive and negative index change, an interpretation that is consistent with Archambault et al. as both can yield, for example, a FBG or a wave-guiding structure. This approach has the benefit of being inclusive of materials that exhibit both positive and negative index change, depending on the laser exposure conditions [11–13]. We believe that the Type I and II classification should be decoupled from attempts to categorize the morphological form of different waveguides, irrespective of the fabrication method employed. In other words, the use of terminology such as Type I or II waveguide should be avoided and replaced with waveguides based on Type I or II modification or similar, the latter already adhered to by some sections of the femtosecond laser direct write community [14–20]. It follows that the proposed Type III classification is inconsistent with the Type I and II classification scheme, and that the standard designations, those originating from the fiber optic community, such as step-index core, graded index core, W-type and depressed cladding, should be used, where possible, to describe different waveguide morphologies.