1. Introduction

There has been many developments of laser processing of materials in recent years. Material sintering, welding, cleaning, machining and cutting [1] using laser has proven efficient and attractive in many fields of applications. Considerable progress in the field of high power laser diodes contributes to making laser processing of materials more efficient and more affordable [2]. However, the optical characteristics of semiconductor lasers limit their uses.

Laser cutting of paper has been widely used in the paper industry. CO₂ lasers are well adapted for this application, both in terms of wavelength, since paper is absorbing at 10.6µm, and in term of price per Watt. Typical cutting speeds are in the range of 150 to 16 000m/s, for typical powers in the 250W to 70kW range [1]. CO₂ lasers are well adapted to industrial applications, and professional cutting systems are becoming available as desktop “plug and play” devices [3]. However, the costs of CO₂ laser systems are not compatible with personal equipment. They require a costly high voltage supply, precise mechanical alignment of mirrors, that prevents their use in low cost low-power desktop applications. Semi-conductor laser diodes are associated with paper processing in laser printing, but in this case the laser is used to write on a photosensitive drum, and not directly on paper.

This paper presents an original approach [4] to perform laser cutting of paper using a single emitter laser diode, within a desktop printer. Optical and thermal properties of paper are first discussed, and our experimental procedure and setup for paper cutting is described. Experimental results are reported and discussed.

2. Experimental details

Our laser cutting system is based on a 1W single emitter diode at 810 nm. The emitter width is 100µm, and the focusing optic consists in two aspherical condensers (L₁: F23 and L₂ : F8). The power available on the paper is determined using a bolometer. The maximum power available is 0.8W, as no special care is taken to minimize parasitic reflections. A beam splitter and a camera can also be used to visualize the focused beam on the paper. The paper is put on a Aluminum drum with black coating. A “low thermal conductivity drum” consisting of a mat of glass fiber was also be used.

Our experiments have been performed on plain white office paper, 80g/m² grade. Several types of inks have been used to investigate the laser cutting. Black office marker has been used. The marking process is slow enough, so that the ink penetrates in the paper and the color begins to appear at the back of the paper. We also formulated an “invisible ink” that absorbs in the Near Infrared from pigment Aldrich IR 820. 7.1mg of dye was dissolved in 4g of water and 1g of propan-1-ol. This ink could be applied using a thin brush or a fountain pen.

 

Fig. 1. Absorbance spectra of white paper, not inked (bottom curve); inked with infrared absorbing ink (middle curve); inked with black marker ink (top curve).

Download Full Size | PDF

Optical properties of white paper inked with the different dyes have been measured with a VARIAN spectrophotometer in the 300 to 1500 nm range. An integration sphere has been used to collect both specular rays and diffusion, to determine the total transmittance T and total reflectance R. The absorbance (A) is deduced from A=1-T-R. Results are presented on figure 1. It can be observed that the white paper absorbs less than 5% of the visible and NIR light. About 80% of the light is reflected and 20% is transmitted. The black marker ink allows 90% or more absorption. The invisible IR ink allows about 80% absorption at the wavelength of the laser, and leaves only a very light pale green appearance. Other pigments are known to absorb NIR light without any visible mark, such as pigment used for soldering plastics using NIR laser diodes [5, 6, 7].

3. Results

The ability of the laser diode to cut the paper marked with black ink is illustrated on Fig. 2, at a laser speed of 1.2 m/min. On Fig. 2a, the laser movement has not been interrupted at the end of the black mark. It can be observed that when the laser continues the path in the area without ink, not even a mark can be seen on paper on both side. Figures 2(b) and 2(c) give closer view of cuts, and it can be seen that the result is tidy on both sides of the paper. Similar experiments have been performed on papers with black marks performed using a laser printer or an inkjet printer. It has been observed that the ink is ablated by the laser, but the paper itself is not cut or even marked. This is easily explained : these inks remain on the very surface of the paper, they are designed for that. This observation is coherent with the use of laser in cleaning superficially stained paper [8] for art conservation, or for generating media with tactile characteristics through ink ablation [9].

Cutting experiments have also been performed on traces made with the invisible NIR ink. A typical result is reported on Fig. 3, corresponding to a cutting speed of 0.75 m/min. The cut is tidy with only a barely visible black line on each side of the cut, corresponding to local burning of the paper.

 

Fig. 2. Laser cuts obtained on paper marked using black marker ink: a) showing interruption of the cut when the laser moves outside the marked area; b) top close view; c) close view on the reverse side of the paper.

Download Full Size | PDF

 

Fig. 3. Laser cut obtained on paper marked using infrared invisible ink.

Download Full Size | PDF

The laser power required to cut the paper has been investigated as a function of the displacement speed, of the spot diameter, and of the ink. Results are reported on Fig. 4, for a spot diameter of 30µm, and for both black marker ink and invisible ink. The cutting threshold is found to increase linearly with the laser power for a large enough power. When the speed is higher or when the beam power is lower, the paper is not completely cut but is partly attacked on its depth. For speeds up to 20% higher than the threshold speed necessary to obtain a complete cut, easily-tearable lines are obtained. For speeds 20% to 100% higher than the threshold speed, easy-folding lines are obtained (Fig. 5).

Further experiments established that the thermal conductivity of the substrate supporting the paper is an important parameter. When a low conductivity support is used, the combustion zone of the paper extends to a width of several mm. The result is not tidy. A similar effect is observed in the case the paper is not properly applied against its metallic support.

 

Fig. 4. Results of laser interaction for different laser powers and speeds, on paper inked using black marker ink (squares) or invisible IR absorbing ink (triangles) ; full symbols indicate a complete cut, and hollow symbols indicate a partial cut. The lines represent the asymptotic dependence of maximal speed cutting speed as a function of laser power.

Download Full Size | PDF

 

Fig. 5. Incomplete cut (scoring) obtained by fast laser movement on paper marked using infrared invisible ink.

Download Full Size | PDF

4. Discussion

For a large enough displacement speed and power, the threshold value of power increases linearly with the displacement speed. This result means that the physical relevant quantity is the energy per unit of length to be cut E_l. This energy depends on the spot width. It can be seen that in our experiments, E_l=0.28J/cm when using black marker ink, and E_l=0.5J/cm when using invisible NIR absorbing ink. It is remarkable that the value we obtain with our 0.8W NIR laser diode system is comparable with that for high power CO₂ lasers [1]: the cutting speed reported with a 250W laser on 75µm thick paper corresponds to E_l=0.9J/cm energy.

A very simple model can be proposed to account for the linear cutting energy E_l. The paper is supposed to degrade at a temperature T_d ; the radial profile of the laser beam is assumed to be uniform over a disk of radius R; the deposition of heat within the paper is supposed to be uniform over a thickness δ (related to the features of optical absorption of the paper at the laser wavelength) ; and heat transfer between the paper and the environment is neglected. Under these circumstances, an order of magnitude of the energy per unit length E_l can be expressed as :

where ρ and Cp are the density and specific heat of paper, and ΔT_d=T_d-T_a is the temperature increase from the ambient temperature Ta to the degradation temperature T_d.

Applying formula (1) with the following numerical values : ρ=800 kg/m³ (corresponding to 80 g/m² paper grade), R=30 µm (value representative of our experiments), δ=paper sheet thickness=100 µm, C_p=1500 J/kg/K (measured by DSC) and ΔT_d=400K (temperature increase at which the paper ignites), one obtains for E_l the value of 0.02 J/cm, i. e. significantly less than the E_l values observed experimentally. The simplistic model described above must therefore be greatly improved in order to better account for the different physical phenomena that occur in the laser paper cutting experiment : optical penetration of the heat deposition, thermal diffusion during and after the laser excitation, and dissipation of enthalpy in the course of the degradation/combustion of the paper. Finally, the presence and nature of the substrate supporting the paper must be taken into consideration in our future thermal diffusion models : in particular, a substrate made of a thermally insulating material will lead to enhanced thermal diffusion in the directions normal to the laser beam, whereas a thermally conducting substrate will pump the heat deposited within the paper sheet. The experimental evidence that paper burns over several mm in the case it is placed on a low conductivity support suggests that two-dimensional thermal propagative effects should be included in a further theoretical approach. It differs from laser ablation processes using high power pulses, where the ablation of the material takes place on a timescale where propagation of the heat can be neglected.

Our process can be easily implemented in inkjet printers, using a special cartridge to dispose the NIR absorbing ink, and a single emitter laser diode [4] to perform the cutting. Inkjet technologies are very mature [10], as well as paper management and positioning. Infrared absorbing dyes are commercially available. High power laser diodes with 100µm emitter width and 5W cw optical power are now commercially available, and 10 W cw diodes with 100µm apertures have been demonstrated [11]. Such diodes are easily fabricated using mass production processes, at low prices for high volumes. In contrast with diodes used for telecommunication systems or pumping, there are no stringent requirements on beam quality or wavelength that limit the available power or increase costs [12]. As their wallplug efficiency may exceed 60%, thermal management is simple and inexpensive.

5. Conclusion

We have demonstrated that paper inked with a suitable absorbing ink can be cut with a laser diode. Invisible infrared ink have been tested and very tidy cuts have been obtained. Cutting speeds up to 1.2m/min have been demonstrated for 0.6W optical power, and this speed scales up with the output power of the diode. Easy-tearable and folding lines can be made by increasing the speed. This low cost and small dimension system is devised to be integrated in inkjet printers, and to offer new opportunities in document creation through cutting and scoring ability [13]. The use of paper at office and at home has increased significantly with the advent of the digital era [14], and these print and cut abilities are expected to contribute to the further development of paper as a popular and versatile tangible medium.