Inferred UV fluence focal-spot profiles from soft x-ray pinhole-camera measurements on OMEGA

The laser-direct-drive inertial confinement fusion (LDD-ICF) concept¹ is used to implode cryogenically layered deuterium–tritium targets on the OMEGA laser.² The implosions produce hot-spot pressures >50 Gbar,^3,4 which is about half of the pressure required to achieve ignition conditions. The goal for the next several years is to demonstrate an ignition-relevant hot-spot pressure of ∼100 Gbar in LDD-ICF. The 100-Gbar project^5–8 includes improvements to the OMEGA Laser System,⁹ diagnostics,^10–13 targets, and modeling,¹⁴ which will lead to a better understanding of the LDD-ICF physics. Simulations with 3-D radiation-hydrodynamic codes suggest that low- and mid-modes in the laser-drive nonuniformity are a major cause of the limitation in the hot-spot pressure.^14,15 The drive uniformity is affected by intensity variations in the 60 UV beams that are overlapped on the target surface. Variations in beam energy, pulse shape, beam pointing, pulse arrival time on the target, and focus size affect the intensity uniformity. This requires a careful monitoring of each beam’s intensity at full laser energy at the target plane. A direct measurement of the intensity uniformity on the target is currently not possible; therefore, it is indirectly inferred from measurements outside the target chamber of the beam energy, the laser power, and the spot size. While energy and power are routinely measured before the final optics assembly for each beam on each shot, the spot size is measured for only a very limited number of beams on some shots. The energy of each beam is measured on each shot at all three harmonic wavelengths with the harmonic energy diagnostic system² before the beams enter the target chamber. The laser pulse shapes are measured on all beams by an optical streak camera system,¹⁶ which provides a measure of the power balance. A timing system measures the arrival time of all 60 beams with picosecond precision and is used to achieve simultaneous on-target arrival time of all beams with an rms variability of 4 ps.¹⁷ The laser beam’s focal-spot profile or far-field fluence distribution is measured outside the target chamber on one beam per shot with an UV equivalent-target-plane (UVETP) diagnostic.¹⁸ Four beams (beams 46, 52, 56, and 57) out of the 60 beams can currently be accessed by the UVETP diagnostic. The question arises whether nonlinear optical effects during the transport of the beam from the pickoff over a distance of about 18 m in air and through the subsequent final optics assembly (phase plate, focus lens, vacuum window, and blast shield) might change the fluence distribution in the target plane compared to that measured by UVETP.

X-ray measurements are routinely performed to obtain information on the beam fluence distribution on the target at full power,^19,20 but this technique is limited in accuracy and dynamic range. It uses multiple pinhole cameras equipped with charge-injection devices to record the time-integrated x-ray emission in the ∼2- to 7-keV photon energy range from a 4-mm-diameter Au-coated sphere that is irradiated with all 60 beams. The technique relies on the knowledge of the relationship between the x-ray fluence F_x and UV fluence F_UV, which are typically connected by a power-law dependence $Fx∼FUVγ.$ An exponent of γ = 3.42 ± 0.13 has been inferred when using a 1-ns laser pulse.¹⁹ This technique has now been further developed to demonstrate an improved accuracy in the measured on-target UV focal-spot profile of selected beams at full power. Using a pinhole camera with a back-thinned charge-coupled device (CCD) camera and softer filtration for x rays with photon energies <2 keV provides a lower exponent of γ ∼ 2 and, therefore, a larger dynamic range in the inferred UV fluence.

The solid curve in Fig. 1 shows a calculated x-ray emission spectrum on a semi-logarithmic plot for the interaction of a 100-ps UV pulse with an Au target. The spectrum was calculated with the 1D radiation hydrodynamic code LILAC²¹ using nonlocal thermal transport, inverse bremsstrahlung laser absorption, and flux-limited multigroup diffusion radiation transport. The equation-of-state model used SESAME tables and the opacity and emissivity table was calculated based on detailed-configuration atomic modeling assuming the collisional-radiative equilibrium limit. The displayed spectrum was obtained from post-processing the LILAC output with the code Spect3D²² using the same opacity/emissivity table as used in the LILAC run, but accessing the full 500-group spectral resolution of the table. The calculation was repeated for various UV fluence values that are applicable to the experiments discussed here. The x-ray spectra were then multiplied with the transmission of a narrow bandpass filter to obtain the x-ray signal F_x at a particular photon energy. The calculations show that the relationship between F_x and F_UV can be described by a power law and that the exponent becomes smaller for lower photon energies. The dashed curve shows the calculated exponent γ vs detected x-ray photon energy on a linear scale. Detection with softer x rays is favorable because for a fixed dynamic range for the x-ray signal, a lower exponent corresponds to an increased dynamic range in the inferred UV signal. It is also favorable with respect to photon statistics because lower-energy photons generate less charge carriers in a CCD pixel. Consequently, a higher number of x-ray photons can be detected until the full well capacity is reached, thereby improving the photon statistics.

In analogy to the UVETP technique, the method described in this paper is called the x-ray target-plane (XTP) method, which is performed at full laser power inside the target chamber. UV fluence profiles were inferred for 100-ps and 1-ns laser pulses for up to 11 beams equipped with SG5-850 distributed phase plates (DPPs)²³ and were compared to directly measured UV profiles from the UVETP diagnostic for 4 out of the 11 beams. Good agreement between the XTP and UVETP measurements was obtained, indicating that nonlinear optical effects from the transport in air and in the optics at the target chamber wall are likely negligible. However, the x-ray measurements are still indirect and limited in the dynamic range. Therefore, a new diagnostic is currently being developed—the full-beam-in-tank (FBIT) diagnostic—which will directly measure the UV fluence distribution and the energy of a single beam in the target chamber after the light has propagated through the final optics assembly at the target plane with full energy and a large dynamic range.²⁴ The beam energy is attenuated by three to six orders of magnitude to be compatible with the signal dynamic range of the CCD detector. FBIT and XTP are complementary, connecting the high accuracy of FBIT for a single beam per shot with the benefit of XTP to characterize multiple beams in a single shot. Comparison of the measured results from the two diagnostic techniques will be used to evaluate any measurement bias.

The experimental setup is schematically depicted in Fig. 2. An OMEGA UV beam propagates from the left to the right side, passes through a distributed polarization rotator,²⁵ and reaches a fused-silica wedge—uncoated on the front and antireflective (AR) coated on the back—that picks up a 4% reflection of the full beam, which is then sent to the UVETP diagnostics. The main beam is then directed over a distance of about 18 m in air and passes through a DPP and a lens that focuses the beam onto a flat foil inside the target chamber. The converging beam passes through a vacuum window and a blast shield. A similar DPP is placed in the UVETP diagnostic directly in front of an OMEGA focusing lens, mimicking the target/beam configuration. The beam is brought through focus in a vacuum tube, which is not shown in the simplified schematic in Fig. 2; outside the tube, the expanding beam is picked up by another lens. The beam is down-collimated and attenuated, and a magnified image of the focus is produced on a CCD camera. Details on the UVETP diagnostic are described in Ref. 18.

The flat-foil target was a 20-μm-thick Si wafer with an area of 6 × 6 mm² that was coated with a 500-nm layer of Au. The target normal was aligned along the axis of an opposing port with a ten-inch manipulator (TIM). Up to five beams were focused simultaneously onto the target such that the laser spots were well separated. A pinhole camera loaded into the TIM imaged the x-ray emission with a magnification of 5.16 onto a back-thinned CCD camera (SI 800). The distance from the target to the pinhole was 46.8 mm. The CCD chip was placed in vacuum, while the electronic part of the camera was located in an air bubble inside the TIM.

The experiment required that the same pinhole be kept for multiple shots and be protected from the target debris. A thin foil (6 μm Al or 19.8 μm Be) was placed 15 mm away from the target as a heat shield. The foil was tilted by about 40° so that the plasma flow was deflected away from the nose cone. In addition, a thin (between 3- and 12-μm) Al foil protected the pinhole from the target debris. The CCD chip was protected with a 12.7-μm Be foil from visible radiation in case the front filter was damaged. The heat shield and the front filter were replaced before each shot and the pinhole was checked. In most of the cases, the pinhole survived and could be used on multiple shots. The survival depends on the total laser energy on the target. The maximum laser energy per beam was limited to ∼40 J for the 100-ps pulse and to ∼450 J for the 1-ns pulse. The beam intensities were in the range of ∼3 × 10¹³ to 1 × 10¹⁴ W/cm² depending on the angle of incidence and the beam energy. Up to five beams per shot were fired when using the 100-ps pulse but only up to two beams could be used for the 1-ns pulse. In addition, a thicker (9- or 12-μm) Al front filter was used for the 1-ns pulse.

Figure 3 shows an example CCD image of the x-ray emission from five laser beams (26, 46, 52, 55, and 61) displayed on a logarithmic scale. A second pinhole provides a replicate image of Beam 46, which is labeled 462. The second images of the other beams fell outside the detection area and were not visible. The distance between the pinholes was known to a high accuracy (better than 1%), providing an in situ measurement of the spatial magnification for each shot. The red rectangles mark the areas that were taken to analyze each beam. The indicated angles are the angles of incidence for each beam with respect to the target normal. With increasing angle, the laser fluence is decreased and the x-ray signal is lower. In addition, various energies were used for the different beams. The beam centers are marked by black crosses. Each laser spot was fitted by a 2D elliptical super-Gaussian function in order to locate the beam center.

The filtration was chosen such that primarily x rays with photon energies <2 keV were detected. Figure 4 shows the calculated spectral sensitivity for the 100-ps interaction using the Al heat shield (red solid curve) and the Be heat shield (green dashed curve). The peaks of both spectra were normalized to unity. The detected spectrum is the product of a calculated emission spectrum (blue dotted curve), the filter transmission, and the quantum efficiency of the CCD. The calculation shows that for the case with the Al heat shield, the detected x rays were in the 1- to 1.5-keV range, while for the case with the Be heat shield, there was a small component of detected x rays in the 2- to 3-keV range.

Each pinhole was optically characterized with a confocal microscope. Figure 2 shows a measured image of a pinhole that was used for the 100-ps pulse experiments. The characterization provided the area of the opening, which was used to estimate the signal before the experiment. Pinholes were selected based on the area of the opening to maximize the dynamic range of the diagnostic while avoiding saturation. Pinholes with a nominal diameter of 20 μm and 8 μm were used for the 100-ps and 1-ns pulses, respectively. The pinhole size also affects the spatial resolution. The point-spread function (PSF) of the pinhole camera was calculated assuming the Fresnel approximation and using the calculated spectral sensitivity shown in Fig. 4. This takes diffraction of the soft x-ray radiation in the pinhole into account. It is important that the same pinhole is kept in for as many shots as possible, including calibration shots, which are discussed in Sec. III.

Effective noise suppression in the analysis is important to obtain high-quality data. The CCD camera always acquired two images per shot: one background image a few seconds before the shot and the data image (foreground) during the shot. The images were significantly affected by high-frequency shot noise. The background image was treated with a median filter over a 9 × 9 pixel area before being subtracted from the foreground image. The resulting image was then treated with a median filter over an area of 3 × 3 pixels. The background and noise were analyzed in a region that was not exposed by the beams (see the white rectangle in Fig. 3) resulting in X_background ≈ 3 analog-to-digital units (ADU) and X_noise ≈ 5 ADU (standard deviation) for this shot. Similar values were obtained for all the other shots.

A histogram analysis of the x-ray images of one beam from two shots with slightly different laser energy allows us to infer the relationship between UV and x-ray fluences. Figure 5 shows an example of such a measurement. The same method has been previously used to measure the nonlinear response of a film.²⁶ The idea is to take the x-ray images of the laser spot of the same beam at slightly different UV laser energies and therefore at slightly different peak fluences. The assumption is that nothing changed in the imaging system, the filtration, and the shape of the spot. The first image of beam 57 in Fig. 5(a) was taken at an energy of E₁ = 32.6 J, which resulted in a significantly brighter x-ray emission than that in the second shot [Fig. 5(b)], which was taken at E₂ = 25.1 J. The histograms and reverse cumulative histograms of both images were calculated [see Figs. 5(c) and 5(d), respectively]. Here, the histograms used 1000 bins. With the known ratio of the beam energies R = E₂/E₁, the reverse cumulative histograms were walked backward to infer the relationship between UV and x-ray fluences [see Fig. 5(e)]. The solid line is a power-law fit to the data providing an exponent of γ = 1.74 ± 0.10.

The black lines in Fig. 5(d) indicate the path that was taken in the retrieval with an arbitrary starting x-ray signal X₁ slightly smaller than the maximum signal X_87299,max in shot 87299. The reverse cumulative histogram (blue solid curve) indicates that about 100 counts belong to a group that had signals between X₁ and X_87299,max. Exactly the same number of counts is reached in shot 87302 (red dashed curve) by a group with signals from X₂ to the maximum signal X_87302.max. Therefore, decreasing the peak UV fluence by a factor equal to R resulted in a decrease of the peak x-ray signal from X₁ to X₂, which provides the first two points in Fig. 5(e). The UV fluence of the first point is normalized to unity and the x-ray signal is equal to X₁. The UV fluence of the second point is equal to R and the x-ray signal is equal to X₂. The value X₂ is then used in the reverse cumulative histogram of shot 87299 to find the next value X₃ in the red dashed curve, and so forth. The UV fluence axis values are given by Rⁱ, where i = 0, 1, …, N − 1 is the step number, and N is the number of steps, in this case N = 15.

The relationship between x-ray fluence and UV fluence has been measured for several shots for the 100-ps and 1-ns square pulses, indicating some fluctuation in the calibration curves (see Fig. 6). The fluctuation is probably the result of shot-to-shot variations in the thickness of the filters, which had to be changed on every shot, photon statistics, and the measurement error of R. The black solid curve shows the average calibration curve and the blue dashed line is a power-law fit to the data, providing γ = 1.77 ± 0.05 for the 100-ps pulse and γ = 2.10 ± 0.06 for the 1-ns pulse. The errors indicate the 95% confidence band from fitting. Both curves were taken with the Al heat shield. It is expected that for the 1-ns pulse, the plasma is hotter and the emitted x-ray spectrum shifted to higher photon energies compared to the spectrum from the 100-ps interaction, which explains the higher γ. A similar measurement with the Be heat shield and the 100-ps pulse yielded γ = 2.19 ± 0.16. This is slightly larger than the value with the Al heat shield and is attributed to the small component of detected 2- to 3-keV x-ray photons when using the Be heat shield (see Fig. 4).

The error of R, which is estimated with ΔR/R = 2.5% for 100-ps pulses and ΔR/R = 0.8% for 1-ns pulses, affects the error of the UV fluence given by ΔUV_i/UV_i = i * ΔR/R. The UV fluence error is a few percent for large UV fluence but becomes increasingly larger at low values, reaching several tens of percent, depending on the number of steps. The error of the x-ray signal is given by photon statistics. The maximum number of photons that can be absorbed in a pixel when reaching the full well capacity is N_ph,sat = X_sat/X_ph ≈ 290, where X_sat ≈ 5 × 10⁴ ADU is the x-ray signal when reaching the full pixel well capacity and X_ph = E_ph/E_e–h/G ≈ 172 ADU is the single-photon signal for a E_ph = 1400-eV photon, which is the average photon energy based on the spectral sensitivity shown in Fig. 4. Here, G = 2.23 electrons/ADU is the gain setting of the CCD and E_e–h = 3.66 eV/electron is the energy required to generate an electron–hole pair in silicon.²⁷ X_sat is slightly below the digitizing limit of the 16-bit camera. The pixel full well capacity of this CCD is ∼1.1 × 10⁵ electrons. The average number of photons in a pixel for a given x-ray signal X can be estimated with N_ph ≈ X (ADU)/172. Based on counting statistics, the associated relative error of a single measurement is given by $∼172/X(ADU)$⁠. The histogram C(X) shows the number of measurements for a particular X value performed, and the empirical standard deviation of the average value is then given by $ΔX/X≈172/XADU⋅CX.$ With typical values of C(1 × 10⁴ ADU) ≈ 100 and C(200 ADU) ≈ 500 [Fig. 5(c)], this results in ΔX/X ≈ 1.3% and ∼4.1%, respectively.

The maximum dynamic range of the x-ray measurement is determined by N_ph,sat because the single-photon detection level is larger than the CCD noise level. The lowest expected single-photon signal is ∼110 ADU for photons with E_ph = 900 eV, which provides a maximum dynamic range of 5 × 10⁴ ADU/110 ADU ≈ 450 in the x-ray signal for an optimal signal level. Figure 4 shows that photons below 900 eV are not detected. The histograms show that there are significant numbers of pixels with signals below 110 ADU, which is caused by events where the absorbed photon energy is split between the adjacent pixels. Those events become visible at very low fluence at the edge of the laser spot. The corresponding maximum dynamic range in the UV is therefore estimated with 450^1/γ ≈ 30 for E_ph = 900 eV and 290^1/γ ≈ 25 for E_ph = 1400 eV, assuming γ = 1.77. This could be improved by detecting softer radiation in the vacuum ultraviolet (VUV) range. For example, decreasing the average detected photon energy by a factor of 4 could increase the UV dynamic range to about three orders of magnitude because 4× more photons can be stored in a pixel to reach the full well capacity and γ ≈ 1 according to the calculation presented in Fig. 1. The difficulty is that this requires very thin and fragile foil filters, which are easily damaged in the harsh plasma environment.

The calibration curves were used to infer the UV profiles. Before that, the x-ray spots were corrected for the ellipticity caused by the angle of incidence of the beam striking the Au foil. Figures 7(a) and 7(b) show an example of the laser spot of beam 56 from XTP and compare it to the directly measured spot from UVETP. For a quantitative comparison of the fluence profiles, both images were fitted with an elliptical 2D super-Gaussian function given by^19,20

where x₀ and y₀ are the coordinates of the beam center, n_SG is the order of the super-Gaussian function, back is a constant background, and a and b are the minor and major axes of the ellipse, respectively. The data analysis is similar to that described in Ref. 20. The fitting is performed over an area of ∼1.6 × 1.6 mm² for both the UVETP and the XTP methods. The minimum signal included in the fit is ∼0.2% of the peak signal for UVETP and ∼2% of the peak signal for XTP. The spot radius is then defined as the geometric mean of a and b,

which describes the average radius, where the fluence is at the 1/e value of the peak fluence F₀. The two main parameters that are used to compare the fluence profiles are R_1/e and n_SG. The degree of ellipticity is given by the ratio of the major axis and the minor axis minus unity. For the XTP data, the fitting process was also done with a convolution of the function of Eq. (1) with the PSF of the pinhole camera. It was found that the PSF had an insignificant effect on the fitting result. R_1/e increased by less than ∼0.1% and n_SG decreased by ∼0.4%, which are smaller than the experimental uncertainties of both quantities.

Figure 7(c) shows the result of the fitting process to the XTP image, while the residual (data minus fit) is shown in Fig. 7(d). The fit parameters were R_1/e = 353.5 ± 0.1 μm and n_SG = 4.86 ± 0.01, where the errors indicate the 95% confidence band from the fitting. The ellipticity was inferred with 0.8%, which means that the beam profile is close to circular. The fitting process of the UVETP data yielded R_1/e = 358.4 ± 0.0 μm, n_SG = 5.03 ± 0.00, and an ellipticity of 1.4%. The fitting values from XTP are slightly lower than those from UVETP. However, this is not significant based on the measurement uncertainties. The statistical errors for R_1/e and n_SG were estimated by repeating the same measurement for the same beam over multiple shots and several campaigns. The results are listed for the 100-ps shots in Table I (UVETP) and Table II (XTP). It was observed that shots with a lower x-ray signal gave significantly different values because of the limitation in the dynamic range. This will be discussed in detail below. For the results in Table II, we averaged only those XTP data that had high peak x-ray signals above 1 × 10⁴ ADU. Table II shows that the statistical errors for R_1/e and n_SG are, on average, ∼1.0% and ∼3.4%, respectively, for the XTP method. These errors are lower for the UVETP method, ∼0.1% and ∼1.9%, respectively, as inferred from Table I.

Causes for systematic errors in the XTP method include magnification errors, calibration errors, and the limitation in the dynamic range. Systematic errors of the direct UVETP method are estimated to be negligible because of a large dynamic range that is more than three orders of magnitude. In addition, differences between the XTP and UVETP results might occur because the same DPP was not used in the UVETP leg and the main beam leg, although for the same DPP design, there might be some slight differences in the two DPPs. This will be addressed in the future by physically swapping the DPPs between the shots.

The magnification error of XTP is less than 1%. The systematic error from the calibration was estimated by using the individual calibration curves shown in Fig. 6(a), instead of the average calibration curve. The errors on R_1/e and n_SG from this effect were estimated with ∼0.5% and ∼0.5%, respectively. The dynamic range affects the inferred values. This is demonstrated in Fig. 8, which shows the inferred R_1/e and n_SG as a function of the peak x-ray signal for Beam 55. With a lower x-ray signal, and therefore, lower dynamic range, a smaller R_1/e and a higher n_SG are inferred. This is expected because with a lower dynamic range, only the peak of the fluence distribution becomes accessible, resulting in a smaller observed spot size and a steeper profile. The solid curves in Fig. 8 show the calculated trend based on an assumed dynamic range of ∼25 for an x-ray signal of 5 × 10⁴ ADU and γ = 1.77. This is in agreement with the previous estimate of the UV dynamic range. The dashed lines in Fig. 8 represent the averaged inferred ⟨R_1/e⟩ and ⟨n_SG⟩. The model indicates that the actual R_1/e value is higher than the inferred average value and that the actual n_SG value is lower. An accurate prediction of the R_1/e and n_SG values for an infinite large dynamic range is not possible because of the large scattering of the data and the limited dynamic range.

The XTP data from the 11 beams with SG5-850 DPP show some spread in the spot size with beam 52 having the largest spot (R_1/e = 365.0 μm) and beam 61 the smallest spot (R_1/e = 345.9 μm). The difference in the spot size between the largest and the smallest beams is 19 μm (∼5%), which is larger than the measurement error; therefore, it is likely to be real. The peak fluence difference on the target between beams 52 and 61 is estimated with ∼11%. The average XTP values over the 11 beams resulted in R_1/e = 356.6 μm and n_SG = 4.92. The average XTP values over those four beams that were covered by UVETP resulted in R_1/e = 358.9 μm and n_SG = 5.03, which actually agree with the averaged UVETP data within the errors (see Table I). Beam 52, however, seems to be an outlier with an R_1/e,XTP, which is 9 μm larger than the R_1/e,UVETP.

Similar measurements were performed for the 1-ns pulses, which covered only four beams so far (see Tables III and IV). The statistical errors are similar to the 100-ps pulse measurement. For the 1-ns measurement, the XTP data of beam 52 are in good agreement with the UVETP data, which might indicate an unaccounted systematic error in the 100-ps XTP measurement of beam 52. The plasma flow and radial heat transport might play a role for the longer pulse duration.²⁸ A lateral gradient in the electron temperature transports heat in the radial direction, which results in a plasma area that is larger than the area irradiated by the beam. In addition, plasma expands from the target surface predominantly along the target–normal direction, but also to some degree in the lateral direction. The plasma expansion velocity can be estimated by the ion sound velocity, which is given by $cs≈106ZTeeV/Acm/s,$ where ⟨Z⟩ is the ionization degree, T_e is the electron temperature in eV, and A is the mass number. For T_e ≈ 1000 eV and ⟨Z⟩/A ≈ 1/2, c_s ≈ 2.2 × 10⁷ cm/s, and the estimated expansion distance l_exp = c_sτ is ∼20 μm and ∼200 μm after τ = 100 ps and τ = 1 ns, respectively. The plasma expansion is primarily 1D along the target normal as long as l_exp is much smaller than the laser spot. For longer times, however, the plasma will also expand radially with a fraction of l_exp, which depends on the radial heat transport. This could lead to a larger inferred laser spot size from the XTP measurements, especially for the 1-ns pulse irradiation. However, the close agreement between the XTP and UVETP data suggest that this effect is not significant.

The variation of the beam spot sizes does have an effect on the peak fluence F₀, which can be calculated using

Here, E_UV is the UV beam energy and a, b, and n_SG are the inferred beam parameters. Assuming the same energy in each beam, the peak fluence was calculated based on the inferred beam parameters. Figure 9 shows the beam-to-beam variation of the peak fluence normalized to the average value for (a) the 100-ps and (b) the 1-ns pulse measurements. The red squares refer to the XTP diagnostic, and the blue diamonds refer to the UVETP diagnostic. The yellow band indicates the acceptable rms variation (σ_rms) in the peak fluence based on variation in the beam shape, which is σ_rms ≈ 2%. The XTP data from the 11 beams indicate σ_rms ≈ 3% for both the 100-ps and the 1-ns pulse measurements, which is slightly larger than the acceptable variation. The UVETP data from the four beams result in σ_rms ≈ 1.5%, which is below the limit.

A method has been developed to infer UV laser fluence profiles on a target at full energy from x-ray measurements in the target chamber. The method is called the x-ray target-plane (XTP) method, and it was simultaneously fielded with direct laser spot measurements from an UV equivalent-target-plane (UVETP) diagnostic outside the target chamber. Measurements were taken for both 100-ps Gaussian and 1-ns square pulses. To infer the UV profiles from the XTP data requires careful calibration measurements of the UV fluence to x-ray fluence, which were taken for both pulse shapes. The calibration curves can be fitted by a power law. The exponent for the 100-ps pulses was determined with γ = 1.77 ± 0.05, while it was γ = 2.10 ± 0.06 for the 1-ns pulse. The maximum dynamic range of the XTP diagnostic is ∼25 for the UV light. This could be significantly improved if the average detected photon energy were lower. It is estimated that a dynamic range of three orders of magnitude should be possible when reducing the average photon energy to ∼300 eV. The UV profiles were fitted by a 2D super-Gaussian function providing two relevant fit parameters: the 1/e radius of the laser spot (R_1/e) and the super-Gaussian order (n_SG). The UVETP diagnostic currently samples only 4 of the 60 beams. XTP data are in good agreement with the UVETP data on these four beams. XTP measurements of 11 beams with the 100-ps pulse show some variation in the spot size that is beyond the measurement error. It is speculated that this might be because of variations in the wavefront aberrations or the beam-smoothing kernel in the different beamlines. The accuracy of the XTP method for inferring R_1/e is estimated with 3.7% and for n_SG with 7.8%. The dynamic range of the current XTP diagnostic is not high enough and the errors are still too large to meet the requirements of the 100-Gbar project. Such a high accuracy will be attempted with a direct measurement in the UV with the newly developed full-beam-in-tank diagnostic.

This material is based upon work supported by the Department of Energy National Nuclear Security Administration under Award No. DE-NA0003856, the University of Rochester, and the New York State Energy Research and Development Authority. This report was prepared as an account of work sponsored by an agency of the U.S. Government. Neither the U.S. Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the U.S. Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the U.S. Government or any agency thereof.