A. Light-current characteristics

We can see (Fig. 2) the light-current (L-I) characteristic (red curve) and voltage current (V-I) of our FP laser measured at 20°C.

Fig. 2:

L-I and V-I characteristics for an AR/HR Fabry-Perot laser diode at 20°C

We obtain a low threshold current around 43mA with a high external differential efficiency of 1W/A. We don’t observe any discontinuous aspect along the curve and we obtain a power of 190mW at 250mA. The blue curve represents the V-I characteristic of the FP laser diode. The turn on voltage has been measured at 1.65 V and the series resistance extracted from the linear regression of this curve is around 1.9Ω.

B. Beam Quality

To calculate the beam quality, or quality factor M², of our FP lasers diodes, we have first measured the near field and far field of the laser then, with the width of both measurements at 1/e², we determinate the M² using (1):

where w_1/e² and θ_1/e² are respectively the half width at 1/e² of the near field and the far field and λ the wavelength.

We can see (Fig. 3) the beam near field at 20°C, 200mA (around 150mW) in the plane of the layer (or slow axis) and the beam far field (Fig. 4) at the same operating condition.

Fig. 3:

Near field of FP laser with AR/HR coating at 20°C, 200mA, P=150mW

Fig. 4:

Far field of FP laser with an AR/HR coating at 20°C, 200mA, P=150mW

We obtain a divergence of 22.5° at 1/e² and we can note the good quality of the single transversal mode. The representation of the near field (Fig. 3) shows a width of 4.8pm at 1/e². From these values, we obtain a good M2 around 1.9 at 150mW. In the fast axis, we obtain a divergence of 68° at 1/e².

These kinds of laser are good candidate to be integrated in an extended cavity.

A. Light-current characteristics

As we can see in this L-I characteristic (Fig. 5), we obtain a DFB laser with a threshold current around 70mA at 25°C with a slope efficiency of 0.37W/A.

Fig. 5:

L-I and V-I characteristics of an AR/HR DFB laser at 25°C

The V-I characteristic shows that the turn on voltage is around 1.63 V and the series resistance of 1.05W. We have obtained the D₂ line of Rb (780nm) at 25°C with an optical power of 25mW. Compare with the FP laser diode, the threshold current of the DFB laser is higher. Two raisons can explain that: first is the fact that the DFB laser is longer than the FP laser (2mm vs 1mm). Second, due to the integration of a grating in the structure, the total losses increase. For these, the slope efficiency of our DFB laser decreases too.

B. Beam Quality

We also measured the beam quality of the DFB laser present previously. Figures 6 and 7 show the result of respectively the near field and far field at the D₂ line of Rb.

Fig. 6:

Near field of an AR/HR DFB laser at the D₂ line of Rb (780nm) at 25°C, 25mW

Fig. 7:

Far field of an AR/HR DFB laser at the D₂ line of Rb (780nm) at 25°C, 25mW

We obtain a width at the facet of our laser of 5pm (at 1/e²) at 25°C and 25mW. The beam divergence in the slow axis is around 22° at 1/e2. The beam quality factor resulting of the calculation is 1.9. We can notice that all these values are the same for DFB and FP laser diodes.

C. Single longitudinal mode at D2 line of Rb

In order to check the single frequency behavior together with the emission at the R_b D2 line (780nm), we measured the optical spectrum (Fig. 8) by coupling the beam with an optical fiber and injected it in an optical spectrum analyzer.

Fig. 8:

Optical spectrum at the D₂ line of Rb for an AR/HR DFB laser at 25°C, 25mW

We obtain the D2 line of Rb for the temperature and current conditions of 25°C and 140mA, the optical power corresponding is 25mW (see Fig. 5). We observed a very good side mode suppression ratio (SMSR) of 48dB.

D. Linewidth characterization

The spectral linewidths of the DFB lasers were measured using the delayed self-heterodyne linewidth measurement method [5]. This method was developed to carry out the determination of two important parameters of semiconductor lasers such as the linewidth and indirectly the Henry factor. The distributed feedback diode laser output is divided into two paths (Fig. 9). The first path is delayed by time t₀ (10.2 ps) through a single mode fiber, 2 km long. The second path is frequency shifted by a frequency f_AOM (80 MHz) that is much higher than the spectral spread to be measured. The throughputs of both paths are mixed by a coupler 50/50 and injected on an avalanche photodiode. When the delay t is much longer than the coherence time of the laser output (factor 5 or 10) the two paths are independent of each other [6]. In these conditions, the relation between the self-heterodyne width at 3dB and the linewidth depends on the kind of noise. The 3dB spectral width is twice the experimental beat value when the spectrum is lorentzian (white noise) and V2 times when it is gaussian (flicker noise or 1/f noise) [5,7].

Fig. 9:

Schema of the self-heterodyne method

We can determinate the relation between the linewidth of the DFB laser and its length of coherence by using the following formula (2):

where Lc is the coherence length, c the light velocity in the fiber and Δv the linewidth of the DFB. It’s mean that for a linewidth of 1MHz, the coherence length is equal to 200 m which gives a delay t around 1 μs. By using a 2km fiber, so a delay time of 10.27μs, we are in the condition explained previously.

Figure 10 shows the beat measurement at the Rb D₂ line (dot curve) with its Lorentzian approximation (red curve).

Fig. 10:

Beat measurement (dot curve) of an AR/HR DFB laser at the D2 line of Rb (25°C, 125mW). The blue curve is a lorentzian fit of the beat measurement.

For the AR/HR DFB presented in this paper, we calculate at the condition for obtaining the D₂ line of Rb, for a white noise approximation, a linewidth of 1.07MHz. If we consider the 1/f noise and we fit this curve with a Gaussian function, we obtain a linewidth around 1.9MHz.