Quantum efficiency, intrinsic emittance, and response time measurements of a titanium nitride photocathode

The photoemission properties of photocathodes affect the ultimate performance of electron accelerator based facilities and the cost. Several major properties of titanium nitride (TiN) irradiated by ultraviolet femtosecond laser pulses were studied. Specifically, the quantum efficiency and intrinsic emittance were measured, and the response time was estimated indirectly. For photon energies higher than the work function, quantum efficiency values of $7\times {10}^{-8}–5\times {10}^{-6}$ were obtained. The measured intrinsic emittance was $0.54\mu \mathrm{m}/\mathrm{mm}$ for the typical wavelength of 266 nm (4.66 eV). When the photon energy was lower than the work function, two-photon photoemission appeared, and the intrinsic emittance increased with the laser intensity. At photon energies of higher than and slightly lower than the work function, the intrinsic emittance ranged from 0.28 to $0.60\mu \mathrm{m}/\mathrm{mm}$ and tended toward a minimum constant of $0.28\mu \mathrm{m}/\mathrm{mm}$ if only single-photon photoemission was considered. The response time of TiN was proven to be close to that of gold by comparing the transverse expansion of electron beams under space charge force for the irradiated photon energy higher than the work function. The autocorrelation measurement of the photoemission of TiN and gold were performed, and a comparison of these results indicated that the two-photon photoemission of TiN excited by a photon energy slightly lower than the work function was prompt.

Figure 1

Schematic of the time delay optical path for an autocorrelation measurement. The incident laser pulse is divided into two pulses with a certain time delay and then are incident to a nonlinear optical crystal or a photocathode. The time delay is adjusted by moving the movable mirror back and forth. The $\lambda /4$ wave plate is used to rotate the polarization of the beam passing through it twice by 90°.

Figure 2

Photon energy dependence of quantum efficiency. The work functions obtained from fitting with the Dowell model [13] and Vecchione monocrystalline model [35] are 4.30 and 4.31 eV, respectively. From the Vecchione polycrystalline model [35] they are 4.26–4.36 eV.

Figure 3

Current density, charge of a single pulse divided by the product of the laser spot area and the pulse duration (FWHM), as a function of laser intensity for the wavelength of 266 nm (4.66 eV) and 310 nm (4.00 eV).

Figure 4

Current density as a function of laser intensity for 310 nm (4.00 eV) in the log-log scale. These data are well fitted by a straight line with a slope of 1.71.

Figure 5

Measured transverse emittance of TiN film as a function of the photon energies. The work functions obtained from fitting with the Dowell model [13] and the Vecchione monocrystalline model [35] are 4.11 and 4.08 eV, respectively; and from the Vecchione polycrystalline model fit [35] the work functions are in the range of 4.00–4.16 eV.

Figure 6

Laser intensity dependence of measured emittance for 310 nm (4.00 eV). The average electron numbers per pulse were about 0.81 for the maximal laser intensity. Fitted with Eq. (7), the emittance is $0.29\mu \mathrm{m}/\mathrm{m}\mathrm{m}$ for single-photon photoemission and $0.96\mu \mathrm{m}/\mathrm{m}\mathrm{m}$ for two-photon photoemission.

Figure 7

Root-mean-square fluorescent spot size for Au (at laser wavelength of 259.66 nm) and TiN (at laser wavelengths of 261.03 and 259.66 nm) as a function of the average number of electrons per pulse. The quantum efficiency of gold is $2.44\times {10}^{-5}$, which is about 5 times that of TiN. To emit the same number of electrons as gold, we need to use 5 times the laser intensity to excite TiN.

Figure 8

Normalized interferometric autocorrelation of excited laser pulses using two-photon photoemission from an Au and TiN photocathode at a wavelength of 310 nm. The normalized photocurrent was the average of the upper and lower envelopes of the interferometric autocorrelation signal and was normalized to the photocurrent when the time delay tends to infinity. The laser intensity of each sub-beam incident on the photocathodes was about $130\mathrm{GW}/{\mathrm{m}}^2$.

Figure 9

Normalized intensity autocorrelation of excited laser pulses using two-photon photoemission from an Au and TiN photocathode at a wavelength of 310 nm. The data was normalized to the photocurrent when the time delay tends to infinity. Because the reflectance of the beam splitter to the two types of polarized laser was different, the laser intensity of s-polarized sub-beam incident on the photocathodes was about $300\mathrm{GW}/{\mathrm{m}}^2$, while p-polarized sub-beam was about $170\mathrm{GW}/{\mathrm{m}}^2$. This led to different photocurrent density, but it did not affect the result of pulse width.

Figure 10

AFM topography image of a TiN photocathode. The rms roughness is about 0.5 nm.

Figure 11

Second-order autocorrelation of an 800 nm 15 fs pulse.