Open Access
Add to BibSonomy
Abstract
Based on our earlier investigations, we continued and intensified our effort on the assessment of laser-induced damage effects in the visible range on a digital micromirror device (DMD) in comparison to different electro-optical imaging sensors such as complementary metal–oxide–semiconductors (CMOS) and charge-coupled devices (CCD). The main two objectives of our current work are: i) to fill the gap for the damage threshold regarding the time scale of picosecond pulses (527 nm) for CCD and CMOS devices and ii) evaluate the performance of a new device, the DMD, with both nanosecond pulses (532 nm) and picosecond pulses (527 nm) and compare the results with those of the CCD/CMOS. In the course of this research, we improved the experimental setup. Furthermore, we characterized the damage caused by laser pulse energies exceeding the laser-induced damage threshold (LIDT). For both the CMOS and CCD cameras, we received damage thresholds of about ${10}\;{\rm mJ/cm^2}$ (picosecond pulses). For the DMD, we obtained LIDT values of ${130}\;{\rm mJ/cm^2}$ (nanosecond laser pulses) and ${1500}\;{\rm mJ/cm^2}$ (picosecond laser pulses). In case of the CMOS devices, we additionally compared the appearance of the damage obtained from the output signal of the camera under test and the microscope images of the surface of the camera. The first visible changes on the surface of the sensor occurred at energy densities that are an order of magnitude higher than the threshold values related to the output signal.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. INTRODUCTION
Studies on the interaction between laser radiation and sensor devices are still of great interest regarding countermeasures (e.g., laser warning receivers) or opto-electronic laser protection measures as well as for the improvement of sensor robustness [1]. One example of opto-electronic devices used for laser protection are spatial light modulators (SLM) [2], in particular digital micromirror devices (DMD). DMDs are, for example, widely used in digital projection display systems and are even on their way into space [3,4]. At our institute, a lot of effort is put into the investigation of laser protection measures that provide wavelength-independent protection [5–8]. One of our concepts to protect electro-optical sensors against laser dazzle is based on the use of a DMD [9]. In this protection concept, a DMD is located at an intermediate focal plane of the imaging sensor in order to protect an imaging sensor behind from laser dazzle. In the course of this work, the following question arose: What would happen if this sensor were exposed to laser irradiation with intensities far beyond the threshold for laser dazzle? Which device will be damaged first—the DMD or the imaging sensor?
There is a plenty of literature regarding the laser-induced damage threshold (LIDT) of bulk materials, thin optical layers, and optical sensors. To name a few works that focus on detector materials, photodiodes, and camera sensors sorted by the pulse length:
- • Burgess et al. used continuous-wave laser radiation of 1070 nm to measure the LIDT of a shortwave infrared (SWIR) camera (focused) and a CCD camera (unfocused). The measured LIDT values are approximately 1.5 and ${6}\;{{\rm W/cm}^2}$ at the objective plane (i.e., in front of the camera lens) for irradiation times 0.5 ms/250 ms for the SWIR camera and 5 ms/250 ms for the CCD camera [10].
- • Kruer et al. investigated laser-induced damage to silicon photodiodes irradiated by laser radiation with a wavelength of 1.06 µm and irradiation times ranging from 17 ns up to 1 s. They reported damage thresholds of ${65}\;{\rm J/cm^2}$ (17 ns), ${74}\;{\rm J/cm^2}$ (0.9 ms), ${233}\;{\rm J/cm^2}$ (21 ms), ${744}\;{\rm J/cm^2}$ (80 ms), and up to ${4100}\;{\rm J/cm^2}$ (450 ms) [11].
- • Becker et al. studied laser-induced damage of CCD sensors by laser pulses of 10 ns at 1064 nm. The first visible changes in the surface of the CCD active area were observed at fluence values of 0.7 to ${0.8}\;{\rm J/cm^2}$ for a 1-on-1 test and 0.4 to ${0.6}\;{\rm J/cm^2}$ for a 10-on-1 test [12].
- • Wang et al. investigated laser-induced damage thresholds at a wavelength of 1064 nm. They obtained exposure limits of ${127.2}\;{\rm J/cm^2}$, ${4.8}\;{\rm J/cm^2}$, and ${0.7}\;{\rm J/cm^2}$ for single-crystal silicon using irradiation times of 1 ms, 10 ns, and 10 ps, respectively [13].
- • Guo et al. investigated the damage effect on CMOS detectors irradiated by laser pulses at 1064 nm. In case of nanosecond laser pulses, the damage threshold was around ${0.38}\;{\rm J/cm^2}$, and in the case of picosecond laser pulses they obtained a threshold value of ${0.02}\;{\rm J/cm^2}$ [14].
- • Cai et al. obtained a value of ${0.156}\;{\rm J/cm^2}$ on Si surfaces of a silicon photoelectric detector with a single femtosecond laser pulse (150 fs) at 800 nm [15].
- • Damage mechanisms to CCD caused by femtosecond laser pulses (100 fs, 800 nm) are reported by Gao et al. They observed dot damage and line damage at a fluence level of ${2.5}\;{\rm mJ/cm^2}$ and ${330}\;{\rm mJ/cm^2}$, respectively [16].
In the literature we could not find any corresponding work on sensor damage in the visible range. However, there are plenty of publications that focus on sensor damage in the near-infrared spectral range. But it was not always possible to find out how the damage threshold was assessed and how it was defined. This makes it difficult to compare damage thresholds in different studies. Often the laser-induced damage thresholds reported in the literature differ very much, which is what triggered us to work on own laser-induced damage measurements.
However, the amount of literature concerning the LIDT of a DMD is very limited. For DMDs, the maximum power density for homogeneous illumination is specified by the manufacturer (Texas Instruments) as ${25}\;{\rm W/cm^2}$ in the visible spectral range [17,18]. Faustov et al. reported a damage threshold above 22 mW for a laser wavelength of 633 nm in the case of laser light being focused onto a single micromirror (${13.7} \times {13.7}\;{\rm\unicode{x00B5}{\rm m}^2}$) [19]. This value corresponds to an irradiance of ${\sim}{12}\;{\rm kW/cm^2}$ and is far above the damage threshold stated by Texas Instruments.
In our previous work, we measured the LIDT of a DMD using a continuous-wave (CW) diode-pumped solid-state laser with a wavelength of 532 nm [20]. At an irradiation time of 0.25 ms, we obtained a LIDT of ${19 - 22}\;{\rm kW/cm^2}$ for the occurrence of visible damage. Furthermore, we measured LIDT values of CMOS and CCD cameras for nanosecond laser pulses as well as for CW laser radiation. Our past and present results for CMOS and CCD are summarized later in the conclusion.
The work presented here is supposed to supplement the results of previous work on LIDT. In the case of the CMOS and CCD cameras we used picosecond laser pulses, whereas for the DMD we used nanosecond and picosecond laser pulses. For readers who would like to delve a little deeper into this topic, we recommend reading the work of Westgate [21], dealing with the determination of LIDTs of 2D imaging arrays. Furthermore, there is a standard procedure for determining the LIDT of optical materials [22–25]. We orientated ourselves on this methodology as close as possible and used the terms and definitions as used in ISO 21254 and ISO 11145 [26]. However, we pursued a different approach to determine the LIDT values of camera sensors: accessing the increase of the damaged area with increasing laser energy instead of calculating the damage probability at each laser energy density. We will get back to this later in this work.
2. SETUP AND EXPERIMENT PROCEDURE
As mentioned above, we conducted investigations on CMOS/CCD cameras as well as on a DMD using a CW laser system at a wavelength of 532 nm. Furthermore, we conducted measurements with pulsed laser radiation by illuminating CMOS and CCD cameras with nanosecond pulses and determining the damage threshold of these sensor systems. For the work presented here, we used similar experimental setups as for the previous experiments [27]. Therefore, we will just give a rough overview here and concentrate on the few changes to the former designs in more detail. Figure 1 shows the schematic design of the experiment.
Fig. 1. Experimental setup: (a) configuration for CCD/CMOS damage testing and (b) configuration for DMD damage testing.
Table 1. Specification of the Devices under Test
The measurement setup comprised two different laser systems covering picosecond and nanosecond laser pulse lengths:
- • Nd:YAG nanosecond pulse laser system (Spitlight Hybrid, InnoLas Laser GmbH): operating wavelength 1064 nm, maximum pulse energy 300 mJ at a repetition rate of 10 Hz and spatial Gaussian energy distribution. We used the second harmonic at a wavelength of 532 nm. The temporal pulse length was 10 ns, and the beam diameter was 6 mm (${1/e^2}$). Changing the delay time between the start of the flashlamp and the opening of the Pockels cell to 190 µs resulted in a reduced pulse energy of 2.6 mJ. This laser system was used to access the DMD damages.
- • Nd:YLF HEHR picosecond pulse laser system (picoRegen IC-440, High ${Q}$ Laser Production GmbH): operating wavelength 1053 nm, maximum pulse energy 2.8 mJ at a repetition rate of 10 Hz. We used the second harmonic at a wavelength of 527 nm at a pulse energy of 1.6 mJ. The beam diameter at the laser exit was 1.4 mm (${1/e^2}$). Using this laser source, we examined both the DMD as well as the CMOS/CCD cameras.
After leaving the laser head, the laser beam passes two folding mirrors. Then a small part of the laser light is split off from the main beam using a pellicle beam splitter and is directed on a silicon photodiode (DET36A/M, Thorlabs Incorp.). This reference photodiode determines the laser pulse energy and is calibrated before each experiment. The main beam is directed to a two-stage attenuation unit. At the first attenuation stage, the laser pulse energy is adjusted roughly by a set of calibrated neutral density filters with different optical densities (ODs) ranging from OD 0.5 up to OD 3.0. The second attenuation stage is a motorized attenuator (Power XP, Altechna) for fine-tuning of the laser pulse energy. It consists of a rotating $\lambda /{2}$ wave plate and a dielectric Brewster polarizer that separates $s$- and $p$-polarized beams so the ratio of the two beams can be adjusted by rotating the wave plate. At last, the laser beam is focused onto the device under test by a lens of $f = {80 mm}$ focal length and $f$-number of $f/{5}{\rm .6}$ (Apo-Rodagon ${N}$ 4.0/80, Qioptiq). To accomplish the most homogeneous illumination of the device under test, we used a ring-shaped light source (HPR2, CCS Inc.), which serves as an inspection light.
In order to achieve comparable LIDT values in units of ${\rm J}/{\rm cm^2}$ according to ISO 32354 and ISO 11145, we have to determine the maximum energy density ${H_{{\rm max}}}$ of the laser spot at the position of the device under test (target plane). To accomplish this, one need to calculate the effective area in the target plane. The effective area in the target plane is the ratio of pulse energy to maximum energy density of the laser pulse. Since we expected a focal spot size of a few tens of micrometers, we used a beam profiler (WinCamD-UCD12 CCD, DataRay Inc., with a pixel size of ${4.65} \times {4.65}\;{\unicode{x00B5}{\rm m}}$) in combination with a microscope lens (M Plan Apo NIR, Mitutoyo Inc., $f = 4\;{\rm mm} $, ${\rm N.A.} = 0.42\;{\rm mm} $) and a tube lens (TTL200-A, Thorlabs, $f = 200\;{\rm mm} $) resulting in a magnification of $M = 50$. In the case of the nanosecond pulsed laser system, we received a spot diameter of ${d_{{\rm ns} =}} = {23}{\rm .2\,\,\unicode{x00B5}{\rm m}}$ (${{1/e}^2}$), and in the case of the picosecond laser system we obtained a value of ${d_{{\rm ns =}}} = {27}{\rm .3\,\,\unicode{x00B5}\,{\rm m}}$ (${{1/e}^2}$). In the case of a focused Gaussian beam, the effective beam diameter is defined as ${d_{{\rm eff}}} = {d_{1/e^2}}/\sqrt 2$. This yields to effective beam diameters of ${d_{{\rm ns,eff =}}} = 16.4\,\,{\unicode{x00B5}{\rm m}}$ and ${d_{{\rm ps,eff =}}} = 19.3\,\,{\unicode{x00B5}{\rm m}}$, respectively. Using these values together with the measured pulse energies, we can now determine the maximum energy densities ${H_{{\rm max}}}$ on the device under test.
In order to ensure comparability of the measurement results to our earlier results, we used cameras of the same type for the current measurements and the same exposure times. The exposure times were chosen in such a way that the pixel value of the camera when irradiated with the annular inspection light corresponds to about half of the maximum pixel value. In the case of the DMD measurements, we used the same device as before, since there were still unexposed micromirror surfaces. All devices are listed in Table 1.
The testing procedure follows the procedure described in detail in our previous publication [20]. We installed the device under test on a 3D translation stage so that we were able to shift the sensor into the focal plane of the lens. To measure the damage threshold, we used the 1-on-1 test mode, which means that each test site was irradiated by a single laser pulse only. To detect any changes in the image of the camera, we observed the output signal of the camera. Before and after each laser illumination, two camera images were taken: 1) an illuminated image with the ring-shaped inspection light switched on and 2) an unilluminated image with a covered imaging sensor to avoid illumination from ambient light. All images of a device under test were taken with the same exposure time (see Table 1). By comparing the images taken before and after laser irradiation of the camera device, we could detect whether laser damage of the imaging sensor occurred or not.
To detect the damage of the DMD, we used an inspection camera (see Fig. 1 VRmFC-22/BW, VRmagic). The inspection camera was moved into the optical beam path to take a picture of the DMD before and after each laser shot. The ring-shaped inspection light was modified in such a way that only light coming from a certain angle reaches the surface of the DMD. It is important to note that the angle between the micromirrors’ normal and the beam axis is ${+}{12}\;{\rm deg}$ in the “on” position and ${-}{12}\;{\rm deg}$ in the “off” position with the consequence that in one position the light was reflected to the inspection camera (“on” position) and in the other it was not (“off” position).
Fig. 2. Sections of images taken with a CMOS and a CCD camera with visible laser damage caused by picosecond pulsed laser radiation. (a) Image acquisition with uniform illumination; (b) image acquisition with an unilluminated sensor.
After each irradiation of the DMD, we took a first image of the DMD in its initial “on” position. Then we switched the mirrors back and forth (“reset procedure”) and took another picture of the same section of the DMD surface. Then the last step was to move the DMD to a new position whether the device was damaged or not and repeat the whole procedure after increasing the laser pulse energy.
In the case of the CMOS camera, we also examined the surface with a light microscope (Axioplan, Carl Zeiss AG) in combination with a color microscope camera (DFC450, Leica Camera AG) via reflected light microscopy in the dark field. By illuminating slightly to the side of the device under test, with the directional reflected light shining past the lens, the smooth flat surfaces facing the lens are shown darkly, while edges and irregularities are shown brightly. The lens (Epiplan Neoflurar ${20\times/0.50}$ HD DIC) together with an eyepiece magnification of ${10\times}$ results in a total magnification of $M = 200$.
3. DATA ANALYSIS AND RESULTS
A. Data Analysis
In our work, pixel damage of imaging detectors is recognized via gray level assessment using two different methods: 1) illumination of the object under test by the inspection light or 2) using the dark image when the sensor is covered. In the case of the DMD, micromirror damage was assessed via the gray level signals of the inspection camera, while the surface of the DMD was illuminated by the inspection light. In the case of color cameras, their images were converted to gray-scale images.
In our previous investigations on laser-induced damage of camera sensors using nanosecond laser pulses, we found that in both the unilluminated and illuminated images the damaged areas show the same shape. This also applies to damage caused by picosecond laser pulses (see Fig. 2). For the data analysis we used the unilluminated images to evaluate the damage threshold of the camera sensor since the postprocessing of these images is less extensive. In the case of the DMD it is obvious that only the illuminated images could be used.
A logarithmic relationship between the damaged area and laser pulse energy is used to determine the damage threshold, which is described in various other publications [16,28–30]. In these investigations, the authors observed in microscopic images the evolution of a series of concentric rings, generated by the exposure of ultrashort laser pulses, which they called a “specific amorphous ring pattern.” Apparently, the ring formation is associated with photothermal damage of the subject due to laser radiation. The spatial irradiance distribution ${\phi _0}$ of Gaussian laser beam is described by
where ${\phi _0}$ is the incident peak irradiance, $r$ is the distance from the beam center, and ${\omega _0}$ is the Gaussian beam radius. Knowing that the irradiance $\phi ({r_{\rm I}})$ at the rim of the damaged area with distance ${r_{\rm I}}$ from the center of the laser spot corresponds to the damage threshold ${\phi _{{\rm th}}}$, we getFig. 3. (a) Section of the microscope image of the color CMOS camera surface and (b) the same section of the output image of the color CMOS camera. The numbers (1) to (5) mark the spots on the sensor that were irradiated with the pulsed laser light. The corresponding laser maximum energy densities are (1) ${2295}\;{\rm mJ/cm^2}$, (2) ${1077}\;{\rm mJ/cm^2}$, (3) ${9.7}\;{\rm mJ/cm^2}$, (4) ${9.5}\;{\rm mJ/cm^2}$, and (5) ${11.4}\;{\rm J/cm^2}$. In the case of the laser maximum energy densities (3)–(5), no noticeable changes in the microscope images were observed.
B. Observation, Morphology, and Threshold of Pulsed Laser-Induced Damage on CMOS Cameras
In the case of exposing a color or monochrome CMOS camera with picosecond laser pulses at low laser maximum energy density, we caused damage limited to a few pixels. The number of pixels showing malfunctions spreads radially with increasing laser maximum energy density until entire pixel lines fail (see Figs. 4 and 5). The blue dashed line in the graphs labeled by the ${H_{{\rm th,fit}}}$ tag marks the threshold of laser maximum energy density after which damage occurs. There is a second dashed line (in black) that marks the threshold above which additional effects occur, leading to a stronger growth rate of damaged pixels. The sections shaded in gray demarcate the laser maximum energy density after which the initial coherent damage structure changes into a scattered pattern and/or line damage (whole lines in horizontal and/or in vertical directions fail) starts to appear. The only data that is not in the gray marked areas is used to perform the fits.
Fig. 4. Size of the damaged area of a color CMOS camera as a function of the laser maximum energy density in a semilog plot. Laser wavelength $\lambda = {527\,\,\rm nm}$; laser pulse duration 8.2 ps. The blue dashed line in the graphs labeled by the ${H_{{\rm th,fit}}}$ tag marks the threshold of laser maximum energy density after which damage occurs. The black dashed line marks the threshold above which additional effects may occur, leading to a stronger growth rate of damaged pixels. The sections shaded in gray demarcate the laser maximum energy density after which the initial coherent damage structure changes into a scattered pattern (light gray) and/or line damage (dark gray) starts to appear. The insets show typical damage patterns in the different energy density regions: (a)$H = {0}{.04\,\,\rm J/cm^2}$, ${28} \times {28}$ pixel; (b) $H = 2.3\,\,{\rm J/cm^2}$, ${40} \times {40}$ pixel; and (c)$H = 100\,\,{\rm J/cm^2}$, ${80} \times {80}$ pixel.
Fig. 5. Size of the damaged areas of a monochrome CMOS camera as a function of the laser maximum energy density in a semilog plot. Laser wavelength $\lambda = {527\,\,\rm nm}$; laser pulse duration 8.2 ps. The blue dashed line in the graphs labeled by the ${H_{{\rm th,fit}}}$ tag marks the threshold of laser maximum energy density after which damage occurs. The black dashed line marks the threshold above which additional effects may occur, leading to a stronger growth rate of damaged pixels. The sections shaded in gray demarcate the laser maximum energy density after which line damage starts to appear. The insets show typical damage patterns in the different energy density regions: (a) $H = {0}{.04\,\,\rm J/cm^2}$, ${28} \times {28}$ pixel; (b)$H = 4.6\,\,{\rm J/cm^2}$, ${36} \times {36}$ pixel; and (c)$H = 16\,\,{\rm J/cm^2}$, ${44} \times {44}$ pixel.
1. Color Camera
Figure 4 shows the results gained by the color camera. The first sensor damage was observed at an energy density level of ${10}\;{\rm mJ/cm^2}$. The result of the curve fitting led to a value of ${H_{{\rm th,fit}}} = \;{9}{.5\,\,\rm mJ/cm^2}$. In the camera image the damage appears predominantly in green colors. Looking at Fig. 4, the value of the second dashed line lies at an energy density level of ${340}\;{\rm mJ/cm^2}$. According to Eq. (2), the slope of the straight line between the two dashed lines is related to the applied laser beam diameter. With this slope we receive a reconstructed beam diameter (rbd) of ${d_{{\rm rdb}}} = {16}{.8}\;{\unicode{x00B5}{\rm m}}$. It represents the required beam diameter on the surface of the test object, if ${H_{{\rm th,fit}}}$ was a physical damage threshold of the sensor surface. In this case, it approximates the measured beam diameter very closely. The slope of the dataset after the dashed line indicates that the disturbances in the camera image at these energy density levels are far bigger than the applied laser beam diameter suggests. Above a value of approximately ${18}\;{\rm J/cm^2}$, line damage starts to appear.
The green line in Fig. 4 represents the fit to the data obtained from the microscope images. It yields an ${H_{{\rm th,fit}}} = \;870\,\,{\rm mJ/cm^2}$.
2. Monochrome Camera
Now let us turn our attention to Fig. 5, which represents the dataset of the examined monochrome CMOS device. We observe a beginning sensor damage starting at an energy density level of ${10}\;{\rm m}\;{\rm J/cm^2}$, which is thus comparable with the observation regarding the color camera. In the images the damage appears as white “hot pixels.” The result of the curve fitting led to a value of ${H_{{\rm th,fit}}} = \;14.2\,\,{\rm mJ/cm^2}$. This threshold is marked with a blue dashed line. With the slope of the fit we receive a rbd of ${d_{{\rm rdb}}} = 18.5\;{\unicode{x00B5}{\rm m}}$. Looking at Fig. 5, the value of the second dashed line lies at an energy density level of ${440}\;{\rm mJ/cm^2}$. The slope of the dataset after the dashed line indicates that the disturbances in the camera image at these energy density levels are far bigger than the applied laser beam diameter suggests. Above a value of approximately ${17}\;{\rm J/cm^2}$, line damage starts to appear.
The green line in Fig. 5 represents the fit to the data obtained from the microscope images. It yields an ${H_{{\rm th,fit,m}}} = \;240\,\,{\rm mJ/cm^2}$.
As a result of the two different threshold analyses, i.e., microscope images and the signal output of the cameras, we find significant gaps between their thresholds. A difficulty that arises from this fact concerns the human eye, where only visible lesions can be detected by an ophthalmologist (this corresponds to our microscope images), but one cannot say how serious the physiological effect of the damage to the vision really is (this corresponds to output images of the cameras).
C. Observation, Morphology, and Threshold of Pulsed Laser-Induced Damage on CCD Cameras
The picosecond laser-induced damage to monochrome and color CCD cameras started as point damage, and, in contrast to CMOS cameras, the damage primary elongated in the vertical direction with increasing energy density (see Figs. 6 and 7). Also in contrast to CMOS cameras, line damage only appears in the vertical direction. In our previous work we already observed this behavior when exposing the CCD camera to nanosecond laser pulses.
Fig. 6. Size of the damaged areas of a color CCD camera as a function of the laser maximum energy density in a semilog plot. Laser wavelength $\lambda = {527\,\,\rm nm}$; laser pulse duration 8.2 ps. The blue dashed line in the graphs labeled by the ${H_{{\rm th,fit}}}$ tag marks the threshold of laser maximum energy density after which damage occurs. The black dashed line marks the threshold above which additional effects may occur, leading to a stronger growth rate of damaged pixels. The sections shaded in gray demarcate the laser maximum energy density after which line damage starts to appear. The insets show typical damage patterns in the different energy density regions: (a) $H = {0}{.016\,\,\rm J/cm^2}$, ${32} \times {32}$ pixel; (b)$H = {0}{.6\,\,\rm J/cm^2}$, ${36} \times {92}$ pixel; and (c)$H = 16\,\,{\rm J/cm^2}$, ${44} \times {112}$ pixel.
Fig. 7. Size of the damaged areas of a monochrome CCD camera as a function of the laser maximum energy density in a semilog plot. Laser wavelength $\lambda = {527\,\,\rm nm}$; laser pulse duration 8.2 ps. The blue dashed line in the graphs labeled by the ${H_{{\rm th,fit}}}$ tag marks the threshold of laser maximum energy density after which damage occurs. The black dashed line marks the threshold above which additional effects may occur, leading to a stronger growth rate of damaged pixels. The inserts show typical damage patterns in the different energy density regions: (a) $H = {0}{.014\,\,\rm J/cm^2}$, ${24} \times {24}$ pixel and (b) $H = {0}{.05\,\,\rm J/cm^2}$, ${24} \times {36}$.
1. Color Camera
Taking a closer look to Fig. 6, we can deduce that the first damage to the sensor started at an energy density level of ${10}\;{\rm mJ/cm^2}$ and is therefore comparable with the damage threshold to the CMOS cameras. The result of the curve fitting led to a value of ${H_{{\rm th,fit,m}}} = \;11.1\,\,{\rm mJ/cm^2}$. This threshold is marked with a blue dashed line. In the images the damage appears as green “hot pixels.” The damage starts to spread vertically from a value of ${400}\;{\rm mJ/cm^2}$. The value of the second dashed line lies at an energy density level of ${170}\;{\rm mJ/cm^2}$. The reconstructed beam diameter ${d_{{\rm rdb}}}$ is approximately 21 µm. Above a value of approximately ${12}\;{\rm J/cm^2}$, full line damage starts to appear.
2. Monochrome Camera
In case of the monochrome CCD camera, the first damage starts at an energy density level of ${8}\;{\rm mJ/cm^2}$ (see Fig. 7). The result of the curve fitting led also to a value of ${H_{{\rm th,fit,m}}} = \;7.9\,{\rm mJ/cm^2}$. The distortion in the image starts to spread vertically from a value of ${35}\;{\rm mJ/cm^2}$. The diameter ${d_{{\rm rdb}}}$ is approximately about 19.7 µm. We did not increase the energy to a value where full line damage emerges because we do not want to destroy the whole sensor in case further investigations on this device should be necessary.
D. Observation, Morphology, and Threshold of Pulsed Nanosecond and Picosecond Laser-Induced Damage on DMD
In general:
As described in the experimental setup, we analyzed the occurrence of laser-induced damaging of the micromirror array using the inspection camera. The change in intensity in the images of the inspection camera observing the DMD may be caused by the following alternatives:
- 1. Decrease of surface reflection (irreversible process)
- 2. Destruction of tilt mechanisms (irreversible process)
- 3. Sticking of micromirrors to the “underlaying landing surface” (reversible or irreversible process)
- 4. Switching from “on” to “off” position by laser light impulse (reversible process)
Note: Micromirror elements that are stuck in the “on” position appear always as bright pixels when illuminated as described in Fig. 1. This behavior has already been described in patents, taking advantage of switching micromirrors by optical impact in case micromirror elements become stuck in a tilted position [33]. They reported the use of pulsed radiation using a wavelength in the range from ultraviolet to infrared and pulse durations of 500 ns or less and pulse energies of 1 mJ or less without damaging the DMD. They described a specific example in which they used single-pulse radiation at a wavelength of 532 nm, pulse duration of 5 ns, and a pulse energy of 1 µJ without damaging the micromirror surface. This pulse energy corresponds to an energy density value of ${0.5}\;{\rm J/cm^2}$ [34].
Since one micromirror element is imaged onto 1.5 pixel of the inspection camera, it can occur that individual camera pixels produce a signal generated by a combination of reflections from damaged micromirrors and undamaged ones.
1. Picosecond LIDT
Looking at Fig. 8, we see a selection of dark pixel patterns resulting from irradiating the DMD for increasing energy laser radiation. In order to distinguish between reversible and irreversible reactions of the micromirrors, we performed the “reset procedure” as described in the experimental setup in Section 3. The corresponding patterns after the “reset procedure” are shown in the middle column of Fig. 8.
Fig. 8. Picosecond laser-induced damage to a DMD. Laser wavelength $\lambda = {527\,\,\rm nm}$; laser pulse duration 8.2 ps. Overview with sections of images of the DMD with increasing laser maximum energy density from top to bottom. At each laser maximum energy density, there is one section with micromirrors switched to optical impact (left row) and one recording of the same section with irreversible damaged micromirrors after the “reset procedure” (middle row). The reset procedure is described in Section 2.
Fig. 9. Picosecond laser-induced damage to a DMD. Laser wavelength $\lambda = {527\,\,\rm nm}$; laser pulse duration 8.2 ps. (a) Number of affected micromirrors as a function of the pulsed laser maximum energy density in a semilog plot. (b) Number of affected micromirrors after the reset procedure as a function of the pulsed laser maximum energy density in a semilog plot.
Figure 9(a) shows the number of micromirrors of the DMD no longer reflecting in the direction of the inspection camera for increasing pulse energy. The number of these nonreflecting micromirrors is calculated from the number of dark pixels of the images of the inspection camera (left row of Fig. 9). The vertical green dashed line marks the threshold from which the inspection camera observes changes in the reflectivity of the micromirrors. The threshold for switching micromirrors is obtained from the intersection of the curve fitting with the $x$ axis and leads to a value of ${H_{{\rm th,fit}}} = \;19.5\,\,{\rm mJ/cm^2}$. It marks the required minimum energy density to switch the micromirrors from the “on” to the “off” position. Figure 9(b) shows the residual numbers of micromirrors of the DMD no longer reflecting in the direction of the inspection camera for increasing pulse energy after the resetting procedure. The blue dashed line marks the second threshold of ${H_{{\rm th,fit}}} = 1.5\,\,{\rm J/cm^2}$ from which irreversible damages take place. The complete failure of entire micromirror lines starts at a value of ${H_{{\rm th,fit}}} = \;6.7\,\,{\rm J/cm^2}$.
2. Nanosecond LIDT
Looking at Fig. 10, we see a selection of dark pixel patterns resulting from irradiating the DMD with increasing energy of pulsed nanosecond laser radiation. In order to distinguish between reversible and irreversible reactions of the micromirrors, we performed the “reset procedure” as described in the experimental setup. The corresponding patterns after the reset procedure are shown in the middle column of Fig. 11.
Fig. 10. Nanosecond laser-induced damage to a DMD. Laser wavelength $\lambda = {532\,\,\rm nm}$; laser pulse duration 10 ns. Overview with sections of images of the DMD with increasing laser maximum energy density. At each laser maximum energy density, there is one section with micromirrors switched due to optical impact (left column) and one recording of the same section with irreversible damaged micromirrors after the reset procedure (middle column). The reset procedure is described in Section 2.
Figure 11(a) shows the number of micromirrors of the DMD no longer reflecting in the direction of the inspection camera for increasing pulse energy. The number of these nonreflecting micromirrors is calculated from the number of dark pixels of the images of the inspection camera (left column of Fig. 10). The vertical green dashed line marks the threshold from which the inspection camera observes changes in the reflectivity of the micromirrors. The threshold for switching micromirrors is obtained from the intersection of the fit curve with the $x$ axis and leads to a value of ${H_{{\rm th,fit}}} = \;2.2\,\,{\rm mJ/cm^2}$. It marks the required minimum energy density to switch the micromirrors from the “on” to the “off” position. This is about an order of magnitude below the value we obtained by irradiation with picosecond laser pulses. Figure 11(b) shows the residual numbers of micromirrors of the DMD no longer reflecting in the direction of the inspection camera for increasing pulse energy after the resetting procedure. The blue dashed line marks the second threshold of ${H_{\rm th,fit}} = \;0.13\,\,{\rm J/cm^2}$ from which irreversible damage takes place. The complete failure of entire micromirror lines starts at a value of $H = \;1.7\;{\rm J/cm^2}$.
Fig. 11. Nanosecond laser-induced damage to a DMD. Laser wavelength $\lambda = {532\,\,\rm nm}$; laser pulse duration 10 ns. (a) Number of affected micromirrors as a function of the pulsed laser maximum energy density in a semilog plot. (b) Number of affected micromirrors after the reset procedure as a function of the pulsed laser maximum energy density in a semilog plot.
4. CONCLUSIONS
We studied the formation of laser-induced damage on CMOS and CCD cameras by means of picosecond pulsed laser radiation. The results of this study are listed in Table 2: the cells highlighted in orange color represent the results of this work and, for comparison, the results of our previous work on LIDT with nanosecond laser pulses.
Table 2. Laser-Induced Damage Threshold for Imaging Sensors Obtained from the 1-on-1 Test Using Pulsed Laser Sourcesa
Comparing the results for CMOS and CCD cameras using picosecond laser pulses, we see that the LIDT values do not differ very much. In the case of nanosecond pulses, the LIDT values also do not differ very much.
We can conclude that for both color and monochrome CMOS cameras irradiated with picosecond laser pulses, the LIDT is up to an order of magnitude lower than for those irradiated with nanosecond laser pulses. In the case of color CCD cameras, the threshold values are 3 times lower for picosecond irradiation than for nanosecond irradiation. For the monochrome CCD camera we have no results to compare with, since the sensor was destroyed at an early stage of the measurements using nanosecond laser pulses.
As already described in our previous work on irradiations with nanosecond laser pulses, we observed “hot pixels” after irradiating the sensors using pulses with energy densities beyond the damage threshold. The same is true for picosecond laser pulses. This may indicate that pulsed laser radiation causes a change in the bandgap of the semiconductor junction or a change in the insulation area, resulting in an increasing leakage current.
Moreover, in the case of irradiating the CCD cameras with energy densities far beyond the damage threshold, we observe damage characteristics extending in the vertical direction. An explanation could be that for most CCD cameras, the readout electronics are arranged in a way that the charge is transferred in the vertical direction along the column to the final row (readout register). To avoid the charges escaping laterally, there are “channel-stops” implanted between the pixels to isolate the charge packets from adjacent columns.
In the case of the CMOS camera, we also compared the LIDT values gained from the output signal of the sensor under test with images of the sensor surface taken by a microscope. The first visible changes on the surface of the sensor occurred at energy densities that are an order of magnitude higher than the threshold values related to the output signal of the camera under test. This observation is an indicator that the characteristics of the semiconductor material and thus the functionality of the sensor changes before visible surface damage is noticeable. As a further result of the two different threshold analysis methods, i.e., microscope images and the signal output of the cameras, we find significant gaps between their thresholds. A difficulty that arises from this fact concerns the human eye, where only visible lesions can be detected but one cannot say how serious the physiological effect of the damage to the vision really is. In future work we plan to take a closer look at this behavior by evaluating the camera output signal and the microscope image of the sensor surface simultaneously.
Table 3. Laser-Induced Damage Threshold for a Digital Micromirror Device Obtained from the 1-on-1 Test Using Pulsed Laser Sources
The thresholds for irreversible damages of the DMD for picosecond and nanosecond laser pulses are higher than those for the CMOS and CCD cameras (see Table 3). Returning to the question posed at the end of the first paragraph, we can conclude that the cameras are damaged much earlier than the DMD. From an economic point of view, it would be preferable if the imaging sensor would be damaged first, because the imaging sensor is much cheaper than the DMD. On the other hand, from the point of view of laser protection, it would be preferable if the imaging sensor would remain undamaged for as long as possible. In the latter case, the DMD acts as a sacrificial element for the sensor system. The system could still be used, whereby one would have to accept corresponding image distortions like color distortions or loss of contrast [35]. The extent of the disturbances depends on the size of the damage that has occurred to the DMD. Thus, the results show that for laser radiation in the investigated pulse range, the DMD cannot be employed as a sacrificial element. Only the threshold value for the switching of the mirrors due to optical impact in the case of irradiation with nanosecond laser pulses is below the damage threshold of the camera sensors. The threshold for switching the micromirrors with nanosecond laser pulses is an order of magnitude below the value we obtained by irradiation with picosecond laser pulses.
Our further work should focus on different laser pulse length (e.g., millisecond laser pulses) and different wavelengths in order to determine the behavior of the LIDT as a function of the wavelength and the pulse length.
Disclosures
The authors declare no conflicts of interest.
Data Availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
REFERENCES
1. M. Schiller, G. Ritt, and B. Schwarz, “Maritime laser measurement system: MARLA,” Proc. SPIE 10795, 107950F (2018). [CrossRef]
2. S. Svensson, S. Björkert, H. Kariis, and C. Lopes, “Countering laser pointer threats to road safety,” Proc. SPIE 6402, 640207 (2006). [CrossRef]
3. A. Travinsky, D. Vorobiev, Z. Ninkov, A. Raisanen, M. A. Quijada, S. A. Smee, J. A. Pellish, T. Schwartz, M. Robberto, S. Heap, D. Conley, C. Benavides, N. Garcia, Z. Bredl, and S. Yllanes, “Evaluation of digital micromirror devices for use in space-based multiobject spectrometer application,” Proc. SPIE 3, 035003 (2017). [CrossRef]
4. M. A. Quijada, A. Travinsky, D. Vorobiev, Z. Ninkov, A. Raisanen, M. Robberto, and S. Heap, “Optical evaluation of digital micromirror devices (DMDs) with UV-grade fused silica, sapphire, and magnesium fluoride windows and long-term reflectance of bare devices,” Proc. SPIE 9912, 99125V (2016). [CrossRef]
5. G. Ritt, M. Koerber, D. Forster, and B. Eberle, “Protection performance evaluation regarding imaging sensors hardened against laser dazzling,” Proc. SPIE 9249, 924908 (2014). [CrossRef]
6. G. Ritt and B. Eberle, “Electro-optical sensor with spatial and spectral filtering capability,” Appl. Opt. 50, 3847–3853 (2011). [CrossRef]
7. G. Ritt, D. Walter, and B. Eberle, “Research on laser protection: an overview of 20 years of activities at Fraunhofer IOSB,” Proc. SPIE 8896, 88960G (2013). [CrossRef]
8. G. Ritt and B. Eberle, “Evaluation of protection measures against laser dazzling for imaging sensors,” Opt. Eng. 56, 033108 (2017). [CrossRef]
9. G. Ritt and B. Eberle, “Electro-optical sensor with automatic suppression of laser dazzling,” Proc. SPIE 8541, 85410P (2012). [CrossRef]
10. C. D. Burgess, C. Westgate, L. Hill, and L. Lewins, “Modelled and experimental laser-induced sensor damage thresholds to continuous wave infrared sources,” Proc. SPIE 10797, 107970T (2018). [CrossRef]
11. M. Kruer, R. Allen, L. Esterowitz, and F. Bartoli, “Laser damage in silicon photodiodes,” Opt. Quantum Electron. 8, 453–458 (1976). [CrossRef]
12. M. F. Becker, C.-Z. Zhang, L. Blarre, and R. M. Walser, “Laser-induced functional damage to silicon CCD sensor arrays,” Proc. SPIE 1624, 67–79 (1992). [CrossRef]
13. X. Wang, Z. H. Shen, J. Lu, and X. W. Ni, “Laser-induced damage threshold of silicon in millisecond, nanosecond, and picosecond regimes,” J. Appl. Phys. 108, 033103 (2010). [CrossRef]
14. F. Guo, R. Zhu, A. Wang, and X. Cheng, “Damage effect on CMOS detector irradiated by single-pulse laser,” Proc. SPIE 8905, 890521 (2013). [CrossRef]
15. Y. Cai, Z. Ma, Z. Zhang, G. Cheng, X. Ye, and D. Cheng, “Experimental study of the damage of silicon photoelectric detector materials induced by repetitively pulsed femtosecond laser,” Proc. SPIE 8796, 87960I (2013). [CrossRef]
16. L. Gao, Z. Zhu, Z. Shao, X. Cheng, and S. Chang, “Electric-induced oxide breakdown of a charge-coupled device under femtosecond laser irradiation,” Appl. Opt. 52, 7524–7529 (2013). [CrossRef]
17. Texas Instruments, “Laser Power Handling for DMDs,” (2012), pp. 1–11.
18. Texas Instruments, “DLP 0.7 XGA 2xLVDS Type A DMD,” (2012), pp. 1–28.
19. A. R. Faustov, M. R. Webb, and D. R. Walt, “Note: toward multiple addressable optical trapping,” Rev. Sci. Instrum. 81, 026109 (2010). [CrossRef]
20. B. Schwarz, G. Ritt, M. Körber, and B. Eberle, “Laser-induced damage threshold of camera sensors and micro-opto-electro-mechanical systems,” Opt. Eng. 56, 034108 (2017). [CrossRef]
21. C. Westgate, How to Determine the Laser-induced Damage Threshold of 2-D Imaging Arrays, SPIE Spotlight (2019), Vol. SL47.
22. “Lasers and laser-related equipment—Test methods for laser-induced damage threshold—Part 1: Definitions and general principles,” ISO 21254-1(E) (2011).
23. “Lasers and laser-related equipment—Test methods for laser-induced damage threshold—Part 2: Threshold determination,” ISO 21254-2 (2011).
24. “Lasers and laser-related equipment—Test methods for laser-induced damage threshold—Part 3: Assurance of laser power (energy) handling capabilities,” ISO 21254-3 (2011).
25. “Lasers and laser-related equipment—Test methods for laser-induced damage threshold—Part 4: Inspection, detection and measurement,” ISO/TR-21254-4 (2011).
26. “Optics and photonics—Lasers and laser-related equipment—Vocabulary and symbols,” ISO 11145(E) (2018).
27. B. Schwarz, G. Ritt, M. Körber, and B. Eberle, “Laser-induced damage threshold of camera sensors and micro-opto-electro-mechanical systems,” Proc. SPIE 9987, 99870D (2016). [CrossRef]
28. G. M. Gusakov, A. A. Komarnitskii, and A. I. Frolov, “Simple method of determining the size of small-diameter Gaussian beams,” Meas. Tech. 30, 762–764 (1987). [CrossRef]
29. J. Jandeleit, G. Urbasch, H. D. Hoffmann, H. G. Treusch, and E. W. Kreutz, “Picosecond laser ablation of thin copper films,” Appl. Phys. A 63, 117–121 (1996). [CrossRef]
30. J. M. Liu, “Simple technique for measurements of pulsed Gaussian-beam spot sizes,” Opt. Lett. 7, 196–198 (1982). [CrossRef]
31. R. M. Wood, “Laser induced damage thresholds and laser safety levels. Do the units of measurement matter?” Opt. Laser Technol. 29, 517–522 (1998). [CrossRef]
32. V. N. Lednev, S. M. Pershin, E. D. Obraztsova, S. I. Kudryashov, and A. F. Bunkin, “Single-shot and single-spot measurement of laser ablation threshold for carbon nanotubes,” J. Phys. D 46, 052002 (2013). [CrossRef]
33. D. J. Weaver, “Repair of digital micromirror device having white defects,” U. S. patent 5,617,242 (1 April 1997).
34. D. J. Weaver, “Unsticking mirror elements of digital micromirror device,” U. S. patent 5,717,513 (10 February 1998).
35. G. Ritt, M. Koerber, D. Forster, and B. Eberle, “Protection performance evaluation regarding imaging sensors hardened against laser dazzling,” Opt. Eng. 54, 053106 (2015). [CrossRef]
