Figure 1

Schematic of the harmonic measurements. Attenuators block the fundamental and second harmonic, allowing measurement of the third harmonic on the direct imager YAG screen. Alternatively, we can isolate the second harmonic by measuring the intensity at the P3S1 YAG screen; the soft x-ray mirrors, upstream of P3S1, absorb the third harmonic. We again rely on the attenuators to block the fundamental.Reuse & Permissions

Figure 2

Schematic of the mirror cutoff measurement. For the soft x-ray line, the ratio of intensities measured on YAG screens P3S1 and P2S gives the transmission for a single mirror, and the total transmission is assumed to be the cube of this ratio. Alternatively, we can compare the intensity on P3S1 (downstream of all mirrors) directly against the intensity at the gas detector (upstream of all mirrors). For the hard x-ray line, attenuators block the fundamental, and we calculate the square of the ratio of third harmonic measured before (P3H) and after (P2H) the second mirror.Reuse & Permissions

Figure 3

Transmission plots show the photon energy cutoffs of the soft x-ray (left plot) and hard x-ray (right plot) experimental beam lines. On the left, blue circles show the cube of the ratio of intensities on P3S1 and P2S, while the red diamonds give the ratio from P3S1 to the full pulse energy. Both results match the expected transmission (dashed green curve) [10, 12]. On the right, blue circles give the square of the ratio of intensities measured on P3H and P2H. The dashed curve gives the expected transmission at the nominal 0.0773° angle; however, we find a better match for 0.083° angle (dot-dashed curve). Hard x-ray measurements rely on the third harmonic, resulting in lower resolution than for the soft x-ray measurements. In both cases, the experimental curves are normalized to one at the maximal value.Reuse & Permissions

Figure 4

Examples of third harmonic measurements for fundamental FEL photon energies of 0.9 keV (left), 6 keV (center), and 8 keV (right). Blue circles show the experimental data, red dotted lines show linear one-parameter fits using the nominal attenuation lengths, while solid green lines show three-parameter fits to both the harmonic content and attenuation lengths. The black dot-dashed line shows the expected drop in intensity for the fundamental only. We typically find around 2% third harmonic for soft x rays, with between 0.2%–2% third harmonic at hard x rays. The 4% level found from the one-parameter fit at 8 keV (right plot) is higher than expected, and may be artificially high due to a lack of measurement points dominated by the fundamental.Reuse & Permissions

Figure 5

Transmission for the zirconium filter as a function of photon energy. The filter is made of $100\mu \mathrm{m}$ of zirconium (dashed blue) and 1 mm of silicon (dotted green), with the combined transmission given in solid red. The zirconium K edge manifests as a sharp drop in transmission at 18 keV.Reuse & Permissions

Figure 6

Plot of measured intensity on the direct imager YAG screen ($y$ axis) vs photon energy ($x$ axis), with the FEL tuned to near 6 keV fundamental and the zirconium attenuator inserted. The dropoff in intensity as the fundamental photon energy crosses 6 keV corresponds to the zirconium K edge at 18 keV; we conclude that the drop in intensity is due to third harmonic content. With the zirconium filter blocking the fundamental, we find that the remaining radiation is almost entirely third harmonic. Electron energy jitter and FEL bandwidth broaden the otherwise sharp K edge seen in Fig. 5.Reuse & Permissions

Figure 7

An example image in the soft x-ray beam line (P3S1) shows the characteristic double lobe structure of the second harmonic. Though gas and solid (beryllium) attenuators strongly suppress the fundamental in this image, a small amount of fundamental radiation remains (Gaussian mode background). Diffraction from the beryllium attenuators produces the uneven speckle pattern.Reuse & Permissions

Figure 8

We measure the harmonic component by varying the gas attenuator strength for 0.9 keV (left) and 1 keV (right) fundamental photon energies. Blue circles show experimental data, red dotted lines show one-parameter fits using the nominal attenuation lengths, and solid green lines show three-parameter fits to both the harmonic content and attenuation lengths. The black dot-dashed line shows the expected attenuation of the fundamental only. We find approximately 0.05%–0.1% second harmonic content.Reuse & Permissions

Figure 9

Energy spread measured as a function of laser heater power. To determine energy spread, we measure the electron beam width in the dispersive region a few meters downstream from the laser heater. By increasing the energy of the laser heater, we are able to triple the energy spread of the electron beam. The energy spread increases as the square root of the laser heater power, except at low laser power where trickle heating causes a noticeable increase in the energy spread [17].Reuse & Permissions

Figure 10

Percentage of harmonic power vs laser heater setting. As the FEL performance drops (dotted red line), the percentage of third harmonic also drops (solid green), indicating enhanced sensitivity to electron beam quality at higher harmonics. The FEL performance was determined by energy loss scans, which measure the energy loss of the electrons due to the FEL process [1].Reuse & Permissions