We performed a series of trials varying laser cooking pattern (increasing density, radius, path length, Δ density and randomness) (Fig. 1a) to characterize energy requirements for laser cooking, ambient cooling effects, cooking pattern efficiency, cooking loss in heated samples (Fig. 1b), color change, and browning capabilities through packaging. Unless otherwise noted, all tests were performed on a single layer of printed chicken (approx. 1.5 mm) (Fig. 1c).
Energy required for food safe temperatures
Achieving food safe temperatures25,26 is vital for qualifying lasers as a processing technique. To assess thermal energy required for cooking, we traced a spiraling trochoidal path (Fig. 1a) with the blue light over the sample.
More passes of the cooking pattern at higher laser speed resulted in much quicker initial temperature increases followed by much slower successive heat increases until food safe temperatures were achieved (Fig.
2a, 2b). Conversely, exposing the sample to a single pass of the laser at a much slower cycle speed decreased the rate at which the maximum temperature in the sample increased.
Triangular prints required 6–8 minutes while square prints required 8–10 minutes of consistent exposure from the blue laser to achieve an acceptable final cook temperature of 70°C. Total laser energy required to heat a printed triangle was approximately 1.8–2.4 kJ and 2.4–3 kJ for a printed square (blue laser operates at 5 W). Based on these energy requirements and the relative size of the two printed shapes, an energy per area of 5–7 MJ/m2 of printed chicken is required to achieve food safe cooking temperatures. A more powerful laser at similar energy flux would accelerate the cook time per unit area of food.
Rate of cooling
The “pulsed heating” effect from laser scanning4 is captured in Fig. 2c and Fig. 3 where each spike occurs when the optical center of the laser passes over a volumetric pixel—or “voxel”—of the printed sample. Because the laser scan gradually spirals towards the center of the sample, voxels along the edge of the sample have higher initial peaks and gradually lose amplitude, while voxels closer to the center of the sample have higher amplitudes at the end of the cooking cycle.
In between each pulse of laser heat, Newton’s law of cooling27 can be visualized (visible in Fig. 2c). The cooling constant k, which is in units of s− 1, determines the exponential rate by which a material cools over time. The average cooling rate for a voxel of chicken was approximately 0.0229 s− 1 with a standard error of 0.0006. This value was computed by taking the average cooling rate across a 30 second span just after a laser pulse for nine different trials. Therefore, under normal ambient conditions, we can approximate the cooling rate as
T(t) = Ta + (T0 – Ta)e− .0229t
where T(t) is the temperature at time t, Ta is the ambient temperature, and T0 is the temperature at time t = 0 s. The rate at which food cools plays an important role in determining the population of microbial pathogens in food28. Therefore, by calculating the cooling rate we can more accurately predict the fluctuations in chicken temperature over time and fine tune laser parameters to ensure the food stays above a certain threshold for food safety and consumption.
Larger discrepancies between real-time and maximum recorded temperature at various time steps (Fig. 4) correspond to less efficient heating patterns since the printed sample experiences greater heat loss to the ambient. Keeping total cook time constant, more ambient cooling is observed with multiple shorter passes (Fig. 4a) versus a single pass of the laser (Fig. 4b – f). With the blue laser pumping continuous energy into the system, maximum temperature continues to increase while real-time temperature follows an increasing oscillatory trajectory where the number of local maximum (peaks) corresponds to the number of cycles (for Fig. 4a) or shells of the trochoidal pattern (for Fig. 4b – f).
Contrary to conventional cooking on a range or in a microwave, which provides more uniform heating29–32, samples that are laser-cooked experience bursts of energy from the moving beam of light. As the laser propagates along the surface of the printed chicken (Fig. 3), a thermal peak follows the path of the beam as it makes contact with the food surface—also evident from Fig. 2c. Factors such as circle density, circle diameter, and speed of the laser all affect the shape and amplitude of the thermal peak.
Cooking loss and color
Weight loss and shrinkage of heated poultry samples are both directly correlated to thermal exposure time (Fig. 5). Dry air heating in an oven at 300°F and 400°F33,34 was used as a comparative study. Oven-broiled samples lost nearly twice as much weight and volume as compared to laser-heated samples for comparable exposure times (Fig. 5a). Shrinkage for laser-broiled samples plateaued around 24% (Fig. 5b) while oven-broiled samples were higher around 40%. Chicken cooked via blue laser showed an increase in L from 70.9 to 89.6, a decrease in a from 7.8 to -0.5, and a decrease in b from 16.3 to 11.2. L increased in oven-broiled samples from 51.3 to 61.1, a decreased from 3.3 to 1.6, and b increased from 16.4 to 22.9. Moreover, cooking the samples with a blue laser did not result in browning on the surface of the chicken—a major reason for them having higher lightness values—while oven-broiled samples showed signs of initial browning on the edges of the crusted samples.
Blue light achieved subsurface cooking to a depth of 2.4–3 mm (Fig. 6a) with no immediate browning to the surface. Samples heated via IR laser (Supplementary Fig. S2, right) had an HAZ nearing 3.7–5 mm deep (Fig. 6b). Variations in cooking depth can be attributed to differences in sample thickness (5–6 mm); thicker samples dissipate more thermal energy to the surrounding raw meat. While the heat-affected zone of the blue laser was smaller, heat from the blue light penetrated deeper into the sample. Conversely, IR light was immediately absorbed by the food and resulted in browning. The larger HAZ, from the sample exposed to the IR laser, can be attributed to heat conduction and higher laser power (IR laser was operating at 8 W and the blue laser was operating at 5 W). Due to the high heat absorption by water of IR light, the beam needed substantial defocusing to prevent the sample from vaporizing (IR laser intensity ~ 5 W/cm2, blue laser intensity ~ 60 W/cm2). Three samples were used to assess cooking depth for the blue laser and six samples were used for the samples cooked via IR laser (Supplementary Fig. S3).
Cooking through packaging
Exposing raw poultry samples—sealed in plastic packaging—to blue and near-infrared light resulted in color changes consistent with protein denaturation35,36 with no surface browning (Supplementary Fig. S4, left). Enclosing the sample reduced moisture evaporation and browning development by limiting oxygen exposure20 (visible in Supplementary Fig. S4, right). Using a trochoidal heating pattern with the NIR laser and leaving an air gap between the plastic enclosure and the top surface of the sample (Fig. 6c), we achieved a brown crusted region along the food surface (Fig. 6d, 6e) resembling a “grill” mark.
Slight browning was achieved on the surface of the sealed chicken sample with the blue laser. Moisture evaporation from the heated food beaded into water droplets along the inside of the bag (Fig. 6c). While both lasers induced protein denaturation (Fig. 6d and Supplementary Fig. S4, left), the NIR laser was more efficient for browning foods through packaging. Slower cooking speeds (approx. 50 mm/min) encouraged heat build up and were most effective for browning with the NIR diode laser.
Spatial resolution of laser cooking
We propose a laser-based cooking approach and calculate the cooking resolution to be approximately 1 mm. To quantify this precision, we exposed a 0.25 in thick sample of chicken to a blue laser (Supplementary Fig. S5), an NIR laser (Supplementary Fig. S6), and an MIR laser (Supplementary Fig. S7). Lines were etched onto the surface of the food and cross-sectional cuts were made to assess the HAZ (see Supplementary Section S1 for more details). Figure 7 shows the resulting linear relationship between the horizontal and vertical depth of the HAZ for the blue and NIR lasers. The MIR laser operated at a power and speed that only resulted in surface cooking with no penetrative heating. The relationship between horizontal (Wlaser) and vertical (Dlaser) HAZ followed
Dblue = 0.247 \(\times\)Wblue + 0.609
Dnir = 0.603 \(\times\)Wnir + 0.311
|for the blue (Eq. 2) and NIR (Eq. 3) lasers, respectively.