Picosecond Laser Ablation of Dentine in Endodontics

A.A. Serafetinides, M.G. Khabbaz, M.I. Makropoulou and A.K. Kar

National Technical University of Athens, Department of Physics, Athens, Greece; 
University of Athens, Faculty of Dentistry,
Department of Endodontics, Goudi, Athens, Greece; 
Heriot-Watt University, Department of Physics, Edinburgh, UK

Abstract. The interaction of picosecond laser radiation with human dental tissue was investigated in this
study, in order to determine the ablation rates and the surface characteristics of the dentine by using scanning electron microscopy (SEM). Dentine ablation was performed by using tooth sections of different thicknesses (0.5–2.0 mm). Dental tissue samples were irradiated in air with the fundamental wavelength and first harmonic of a regenerative amplifier Nd:YAG laser system, at 1064 nm and 532 nm, respectively, with a pulse duration of 100 ps and a pulse repetition rate of 10 Hz. The results showed very clean craters surrounded by minimum melting of the surface of dentine when the 1064 nm pulses were used. In contrast, when the first harmonic 532 nm pulses were used, the SEM examinations revealed cracks and melting of dentine with irregular surface modification. Consequently, it seems that cleaning and shaping of the root canal walls during endodontic therapy with the picosecond Nd:YAG laser application may be possible in the future. The, as yet unexplored, field of the picosecond laser interaction with hard dental tissue is expected to be a potential alternative for powerful laser processing of biomedical structures.
Keywords: Endodontics; Laser ablation; Nd:YAG laser; Picosecond laser; Root canal therapy

INTRODUCTION
Lasers were introduced into dental research in the early 1960s but they are still limited in their ability to remove sound tooth structure efficiently and safely. Early results [1] showed that lasers cannot be used in restorative treatment because the energy required for dentine ablation would cause extensive coagulative
necrosis of the pulp tissue. Later investigations indicated that even under low energy density irradiant conditions, slight surface melting could be detected [2], which indicates very high temperatures [3,4]. Bahall et al. [5] confirmed that cement lysis, bone resorption and ankylosis occurred as a result of thermal
injury of the periradicular tissue after the use of the Nd:YAG laser treatment in the root canal of animal teeth. Heat diffusion plays a very important role in photothermal laser– tissue interaction mechanisms.

Using pulse durations in the microsecond and nanosecond ranges, the expansion of hot plasma primarily results in a shock wave ablating the surrounding tissues. Pulsed Er:YAG and HF lasers have attracted much attention because they operate at a wavelength (λ=2.94 m) where hard
tissues, by virtue of their water and hydroxyapatatite content, exhibit strong absorption restricting residual thermal damage to a
relatively small zone [6,7].


At present, the use of laser light in dentistry (mainly CO2, Nd:YAG, Ho:YAG and argon lasers), is accepted for the management of soft tissue in clinical practice (e.g. in oral and maxillofacial surgery) and in the polymerisation of light-activated restorative materials, whereas other dentistry specialities, such as endodontics, are still waiting for a safe and efficient laser procedure [8].


The ablation of dental substances using ultrashort, picosecond laser pulses was proposed as an alternative to longer laser pulses because the energy threshold for ablating biological tissues varies approximately with the square root of pulse duration [9].


Niemz [10] reported picosecond laser ablation of tooth enamel, using a Nd:YLF oscillator laser at 1053 nm wavelength in 1994. In one of our previous studies [11] we applied different types of lasers with picosecond and femtosecond pulse durations to dentine.


Despite the ultrashort pulse duration, our results indicated good cutting e$cacy, comparable to or better than other nanosecond lasers without any macroscopically observed thermal damage of the surrounding tissues.

AIM OF THE STUDY
The aim of the present investigation was to study the quantitative (ablation rates) and the qualitative results (surface characterisation) of the interaction of dentine with the 100 ps Nd:YAG laser pulses at 1064 nm and 532 nm.


MATERIALS AND METHODS
A total of 30 freshly extracted human teeth were longitudinally sectioned under running water at various thicknesses ranging from 0.5 to 2 mm, and kept in water at room temperature. Before the start of the laser experiments, the dental sections were examined by light microscopy for detection of cracks. Any cracked sections were discarded. The dental sections were then positioned vertically in a holder. To avoid ablation debris deposition, the laser pulses were fired horizontally onto the dentine surface, regardless of the anatomical area of the tooth.


Dental tissue sample targets, three for each laser fluence, were irradiated in air with different irradiation conditions, as follows:
1. A regenerative amplifier system (RGA) provided laser pulses at 1064 nm wavelength (Nd:YAG laser amplifier Coherent Antares),
100 ps pulse duration and 10 Hz pulse repetition rate. The pulse duration was observed on a 769J Tektronix main frame scope which incorporates a 7512/TDR sampler, and laser energy was measured with a Rj 7200 energy meter. The laser beam was focused onto the targets using a 10 cm focal length BK7 lens.
2. The first harmonic from the RGA laser output provided laser pulses at 532 nm wavelength, with pulse duration 100 ps and pulse repetition rate of 10 Hz. The experimental set up is presented schematically in Fig. 1.


For energy fluence calculations the beam spot size was determined by replacing the hard tissue samples with pieces of developed Polaroid film and the etched spot was measured by an accurate travelling microscope having an ocular calibrated to 10 m.
Different laser energy values were obtained by filtering the maximum laser energy output with a set of special neutral density filters. The transmission figures of the filters were used to calibrate the laser energy output values used in the perforation experiments. The energy
fluence at each setting was calculated as the ratio of the laser energy per laser spot area. 


For all experiments, fluence measurements are accurate within ±10%. 

The end point of laser ablation of the samples was perforation of the tooth. The ablation rate per pulse at different energy fluence settings was calculated by measuring the time needed for the perforation of the measured whole dental sample thickness.

Postablative surface characteristics (degree of charring, cracks and other surface deformation) were evaluated using light microscopy
(KAPS D 35614) and scanning electron microscopy (SEM; JEOL JSM 6100).

The SEM images presented in this paper correspond to the following laser parameters. For the fundamental wavelength Nd:YAG
laser pulses the laser energy was 57 mJ and the spot size 0.985 mm², resulting in an energy fluence of 5.786 J/cm². For the first harmonic
Nd:YAG laser pulses the laser energy was 28.1 mJ and the spot size 0.882 mm², resulting in an energy fluence of 3.186 J/cm².


RESULTS
Quantitative Evaluation
For quantitative evaluation of the ablational behaviour of the picosecond laser sources used in this work, dentine ablation rates versus
laser energy fluence were plotted.


The small number of dental samples irradiated for each fluence and the fact that the irradiated areas of the dental sections are not
identical (root and crown), do not allow us to add statistically significant error bars to the figures. Despite this, the maximum possible
error in our energy fluence measurements is estimated to be in the order of
±10%. Another source of variation in ablation rate reproducibility is the great inhomogeneity within individual samples and between different samples (e.g. differences in tissue density, microstructure, water content, regional thickness, mechanical strength).

Results show that the ablation rate of the infrared Nd:YAG laser pulses was low (Fig. 2), whereas the ablation rate of the visible picosecond laser pulses increases rapidly and it can be fitted with a sigmoidal fitting curve as shown in Fig. 3.


Qualitative Evaluation
Macroscopic Observations

No charring was observed with either of the wo laser wavelengths at any of the pulse energies tested.
SEM Observations
The surface morphology images of the interaction of the Nd:YAG laser with dentine are shown in Figs 4–6. It is obvious from these figures that the surfaces of irradiated dentine are very clean without any smear layer andwith opened tubules. Figure 4 shows a view of a conical-shaped crater with irregular edges. 

This irregularity probably reflects tissue inhomogeneities, in combination with reduced laser beam intensity in the periphery of the spot area. No cracks are detected. A detail from the sample of Fig. 4, with greater magnification, is shown in Fig. 5. The irregular edges are obvious.


Melted materials are observed away from the edge (right upper corner of the image), and dentine tubules are opened near to the border of the crater. Also a smear layer is not detected in contact with the edge of the crater. Small grained dentine in the border of the crater with some dentine chips is presented in Fig. 6. The surface morphology images of the interaction of the first harmonic Nd:YAG laser with dentine are shown in Figs 7–9. The surfaces of irradiated dentine are melted and the tubules are sealed off. Cracks are observed. Figure 7 shows a view of a crater surrounded by irregular and rough edges; intensive cracking is evident. Dentine surrounding the crater is coated with a smear layer.

Figure 8 shows normal and irradiated dentine. Normal dentine is totally coated with a smear layer. This is evidence of the absence of thermal damage to surrounding tissues. The surface of irradiated dentine is composed of granular melted material spread uniformly throughout the surface. The edge of the crater is extremely sharp.

A magnification of the irradiated area, which is composed of finer beads of melted material, is illustrated in Fig. 9. There is selective ablation of intertubular dentine and the tubules are sealed off with fused and grained dentine.

DISCUSSION
It is well established that cleaning and shaping of the root canal system is an essential objective for the success of endodontic therapy. The
aim of this procedure is to eliminate all tissue remnants as well as bacteria from the root canal system. Shaping creates smooth wall surfaces of root canals and suitably shaped space that facilitates filling [12].

In a previous study we found that dentine ablation rates with near-infrared and visible picosecond laser pulses are comparable to or better than those of the excimer nanosecond lasers [11], but approximately one order of magnitude less e$cient than those of the mid-infrared HF laser [6].
Comparing the cutting efficiency of the Nd:YAG and the first harmonic Nd:YAG laser in this study, we observed that the latter creates a crater of 2 m in dentine with an energy fluence of 0.6 J/cm² (Fig. 3) whereas the Nd:YAG laser needs a higher fluence of 4.5 J/cm² to create the same crater depth (Fig. 2).
Despite the higher energy level needed for the ablation of dentine with the Nd:YAG laser, this ablation seems to create no other mechanical or thermal damage in this hard dental tissue, as revealed by SEM examinations of specimens. This result provides evidence that the ablation of hard dental tissue does not depend entirely on the energy level of the laser used; other parameters need to be considered, such as the laser wavelength and the optomechanical properties of the tissue. 
Macroscopic inspection of the ablation crater showed smooth surfaces, free of thermal damage, for both laser wavelengths used.
Detailed examination of SEM images showing the interaction of the picosecond Nd:YAG laser (λ=1064 nm) with dentine showed a  homogeneous surface with very small grains and a small increase in tubule diameter which means very little rise in the temperature of the
dentine surface with a very small degree of melting. Willms et al. [13] investigated the suitability of picosecond Nd:YLF laser, emitting at a wavelength of 1053 nm, for dentine ablation and also reported no thermal or mechanical damage.

Correlation of the pulse duration with the cutting quality of the near-infrared laser radiation requires comparison of our results with the results obtained by other researchers with longer laser pulses. Neev et al. [14] investigated the ablation e$ciency, thermal damage and surface  characteristics of dentine irradiated by the Nd:YAG laser with a pulse duration of 15 ns and found that the surface of the dentine had a glazed appearance with crack formation when an 80 mJ/pulse at 10 Hz pulse repetition rate was used. Ragot-Roy et al. [15] using the Nd:YAG laser, but with a pulse duration of 65 s, found that laser irradiation of root canal dentine surfaces induces a nonhomogeneously modified dentine layer, where melted and resolidified dentine partially closes the dentine tubules.
In an attempt to explain the basic mechanisms of ultrashort laser ablation of hard tissues, we have considered the following. The Nd:YAG laser operates at a wavelength of 1064 nm. This wavelength corresponds to an energy per single photon emission of only 1.18 eV, which is not enough to ionise tissue molecules and to break molecular bonds. Multiphoton absorption is suppressed because of the low absorption coe$cient of the tooth tissue at this wavelength (the range of 300–1300 nm is a range of low absorption for enamel and dentine [7]). However, the absorption can be enhanced and a multiphoton process can be initiated in so-called ‘plasma-mediated ablation’ [9]. Due to the high power densities per pulse (>1012 W/cm²), a multiphoton process can take place, providing initial electrons for the optical breakdown. The strong field in the laser beam focus (>106 V/cm), leads to an avalanche-like multiplication of electrons. The generated plasma is further heated by inverse
bremsstrahlung resulting in an enhanced shielding of the tissue sample. As the absorption of successive laser pulses is strongly enhanced by the plasma, interaction effects are mainly localised at the surface of the target [10]. The rapid expansion of this plasma generates an acoustic shock wave which finally enables ablation of the tissue. Therefore, in contrast to longer pulse duration, picosecond pulses at near-infrared laser wavelengths ablate tissues with no signs of thermal damage, as heat diffusion into the tissue is negligible and the energy loss into the target is minimised [10,13]. 


The interaction of visible ultrashort pulsed laser radiation with dentine was also investigated, in our study, by using the first harmonic of the Nd:YAG laser. 


As shown in Fig. 1, the first harmonic of the Nd:YAG laser in our system is produced by passing an Nd:YAG incident beam of 1064 nm through an integral BBO harmonic generator to produce an intense visible green laser light of 532 nm, with a pulse width of 100 ps. This wavelength enhances the energy absorption of blood-perfused tissues and avoids the requirement of a separate visible aiming beam [15].


In our study, ablation of dentine by the first harmonic Nd:YAG was associated with intensive cracks and melting while dentine tubules were closed. The melting of dentine demonstrates that in conjunction with the primary photomechanical mechanism secondary thermal effects are also observed. The sigmoid nature of the ablative response with increasing energy density and the conical nature of the craters produced is a result of increased laser beam absorption in the plume of ablation produced at higher fluences [16]. Tewfik et al. [17] reported on the  structural and functional changes in root dentine following exposure to KTP/532 laser, with laser energy per pulse of 1–2 J and pulse lengths of 0.2–1 s. They also found loss of the smear layer, smear plugs, and peritubular dentine, thus increasing the tubular diameter, after 10 pulses with approximately 1 min between pulses to allow for temperature dissipation. Machida et al. [18] irradiated dentine samples with di#erent laser power parameters, but with the same total laser energy, 6 J, by using the KTP:YAG laser at λ=532 nm. They reported no significant
microstructural changes of dentine when the total laser energy of 6 J was delivered with 1 W6 s, repeated five times and with 2 W3 s, repeated five times. However, after irradiation with 3 W2 s, repeated five times, SEM examination demonstrated successful removal of smear layer and debris with a few localised patches of melting. The comparison of the surface morphology of visible and near-infrared laser ablated craters in dentine indicate that di#erent ablation mechanisms are involved. In both cases, efficient ablation was observed only in the presence of plasma. However, the ablation mechanism in the case of the fundamental wavelength Nd:YAG laser could be better described as ‘plasma-mediated ablation’, whereas in the case of the first harmonic Nd:YAG laser the observed mechanical disruption indicates that the relevant mechanism could be better characterised as ‘photodisruption’. According to Niemz [10], if the ionising effect of plasma contributes more to the ablation process than the generated acoustic shock wave, the mechanism is called ‘plasma-mediated ablation’. If mechanical disruption is induced as a primary result of acoustic shock waves, the mechanism is characterised as ‘photodisruption’.


CONCLUSIONS
The macroscopic inspection of the ablated dentine in the present study shows smooth surfaces, free of thermal damage. SEM examination in the case of Nd:YAG laser, demonstrated successful ablation of dentine and removal of smear layer and debris with opened orifices. These successful properties achieved using this technique should significantly improve root canal preparation and seal and tooth longevity in a clinical situation. In contrast, the use of the first harmonic Nd:YAG laser provided an irregular surface of dentine and cracks. Consequently, it seems that the cleaning and shaping of the root canal with the application of the picosecond Nd:YAG laser may be possible and research needs to be continued.


Dentine ablation measurements with nearinfrared and visible picosecond laser pulses showed a relatively good effciency. In all systems, efficient ablation was observed only in the presence of plasma. A ‘plasma-mediated ablation’ mechanism was considered for the dentine ablation with 1064 nm laser pulses, whereas the ‘photodisruption’ process was involved mainly in the dentine irradiation with 532 nm laser pulses.


Despite the relatively slow laser efficiency for the dentine ablation rates, ultrashort-pulse laser ablation of hard dental tissues is opening up a previously unexplored field which is expected to impact on laser processing in dentistry. These investigations can exhibit new insights related to the versatility of laser surface processing in general.


ACKNOWLEDGEMENTS
The authors would like to thank the members of the Non-Linear Optics Group of the Heriot-Watt University for their help in carrying out the ablation experiments.


REFERENCES
1. Adrian J, Berniet J, Sprague W. Laser and the dental pulp. J Am Dent Assoc 1971;83:113–17.
2. Boehm R, Rich J, Webster J et al. Thermal stress effects and surface cracking associated with laser use on human teeth. J Biomech Eng 1977;99:189–94.
3. Kuroda S, Fowler B. Compositional, structural, and phase changes in vitro laser-irradiated human tooth enamel. Calcif Tissue Int 1984;36:361–9.
4. Lobene R, Bhussry B, Fine S. Interaction of carbon dioxide laser radiation with enamel and dentin. J Dent Res 1968;47:311–17.
5. Bahall J, Howard P, Miserendino L. Preliminary investigation of the histological e#ects of laser endodontic treatment on the periradicular tissue in dogs. J Endodont 1992;18:47–51.
6. Makropoulou M, Serafetinides A.A, Khabbaz M et al. Dentin ablation measurements in endodontics with HF and CO2 laser radiation, Proceedings of Medical Applications of Lasers III, 1996;SPIE 2623:200–10.
7. Altshuler GB. Human tooth in low and high intensive light fields. Proceedings of Medical Applications of Lasers III, 1996;SPIE 2623:68–87.
Picosecond Laser Ablation of Dentine in Endodontics 173
8. Wigdor H, Walsh J, Featherstone J et al. Lasers in dentistry. Lasers Surg Med 1995;16:103–33.
9. Stern D, Puliafito CA, Dobi ET, Reidy WT. Corneal ablation by nanosecond, picosecond and femtosecond lasers at 532 nm and 625 nm. Arch Ophthalmol 1989; 107:587–92.
10. Niemz M. Investigation and spectral analysis of the plasma-induced ablation mechanism of dental hydroxyapatite. Appl Phys 1994;B58:273–81.
11. Serafetinides AA, Makropoulou M, Kar A et al. Picosecond and femtosecond laser ablation of hard tissues. Proceedings of Laser Applications in Medicine and Dentistry. 1997;SPIE 2922:200–8.
12. Walton R, Torabinejad M. Principles and Practice of Endodontics, 2nd edn. Philadelphia: Saunders, 1996:201.
13. Willms L, Herschel A, Niemz M et al. Preparation of dental hard tissue with picosecond laser pulses. Lasers Med Sci 1996;11:45–51.
14. Neev J, Pham K, Lee J et al. Dentin ablation with three infrared lasers. Lasers Surg Med 1996;18:121–8. 

15. Ragot-Roy B, Severin C, Maquin M. Pulsed Nd:YAG laser in endodontics. Proceedings of Medical Applications of Lasers II 1995;SPIE 2327:135–43.
16. Bowker TJ, Cross FW, Rumsby PT et al. Excimer laser angioplasty. Quantitative comparison in vitro of three ultraviolet wavelengths on time ablation and haemolysis. Lasers Med Sci 1986;1:91–9.
17. Tewfik H, Pashley D, Horner J et al. Structural and functional changes in root dentin following exposure to KTP/532 laser. J Endodont 1993;19:492–7.
18. Machida T, Wilder-Smith P, Arrastia A et al. Root canal preparation using the second harmonic KTP:YAG laser: a thermographic and scanning electron microscopic study. J Endodont 1995;21:88–91.