In our previously published paper [1], the repetition rate of the 90 fs laser system was reported to be 80 MHz but is 40 MHz instead. In the following we correct all derived values, such as pulse energies, peak powers and slope efficiencies which are related to that particular laser system.

In Fig. 3 of the original paper [1] (here Fig. 1), the pulse energy and peak power axes of panels b and e are scaled by a factor of two as shown above in the corrected figure.

 

Fig. 1. Original Fig. 3 in [1]. Measured spectral evolution of third harmonic (a, b, c) and pump (d, e, f) for different input pulse durations (left: 30 fs, center: 90 fs, right: 850 fs). Insets in top row: observed HE₁₃ mode. For all plots: left axes – near-infrared pulse energy; right axes – near-infrared peak power. Molecular fraction is stated in bottom row. Color scale in f applies to bottom row. Marker in f denotes onset of harmonic generation.

Download Full Size | PDF

 

Fig. 2. Original Fig. 4 in [1]. Experimentally measured harmonic signals for increasing pump peak power (a) and pump pulse energy (b). The legend refers to the different pulse lengths and applies to both panels. The upper limits of the y-axes in (a) and (b) are identical.

Download Full Size | PDF

For the 90 fs pulses, in Fig. 3(b) of the original paper [1] (here Fig. 1(b)) the onset peak power is now slightly increased above 3 kW but is still similar for all pulse durations (corrected figure above). The ratio at which the harmonic peak shifts for 90 fs pulses for increasing pulse energy and peak power is reduced by a factor of two and amounts to $R_P^{90} = - 29\textrm{pm}/\textrm{kW}$ and $R_E^{90} = - 323\textrm{pm}/\textrm{nJ}$, where the subscript denotes peak power (P) or pulse energy (E), respectively. The corresponding values for 30 fs remain unchanged: $R_P^{30} = - 14\textrm{pm}/\textrm{kW}$ and $R_E^{30} = - 456\textrm{pm}/\textrm{nJ}$.

Further, the slope efficiencies and photon yield given in Table 1 are changed as follows:

Table 1. Summary of benchmark figures / conversion efficiencies for THG in TCE-LCF

View Table

Compared to the original manuscript, the photon yield is enhanced by a factor of 2, whereas the slope efficiencies versus pulse energy and peak power are reduced by a factor of four for 90 fs pulses. The slope efficiencies now continuously increase (decrease) for longer (shorter) pulses versus pulse energy (peak power).

Figure 4 (here Fig. 2) is corrected as shown above. The previous qualitative discussion of the resulting slope efficiencies remains unchanged, slope efficiencies for shorter pulses still exceed the one for 850 fs pulses, yet the photon yield is still greater for the longer pulses.