Self-Injection Locking of a Distributed Feedback Laser Diode Using a High-Finesse Fabry-Perot Microcavity
Round 1
Reviewer 1 Report
The authors investigated the stabilization of a DFB laser by a Fabry-Perot Microcavity experimentally. Although the spectral linewidth achieved is not particular spectacular, the paper describing the experimental approach is worth to be published if it is revised properly.
In the Introduction, the authors should also cite some papers on the stabilization of DFB lasers by external cavities, such as Liang et al. Nature Communications DOI: 10.1038/ncomms8371, Lewoczko-Adamczyk et al OPTICS EXPRESS DOI:10.1364/OE.23.009705, and others. How the theoretical linewidth of the FP cavity was estimated? Explain “Q factor 1.6166×104”. Describe the DFB laser (cavity length, type of grating, coupling coefficient, facet reflectivities etc.) Which electrical driver of the DFB laser has been used? What does it mean: “same polarization of the DFB laser” “It can be seen that the output wavelength of the free-running laser significantly goes up as the temperature of laser increases (black line) due the temperature dependence of cavity length.” There are two effects affecting the laser wavelength: the temperature dependencies of the effective refractive index and the cavity length. Estimate the contributions of both effects. “62 mV output power level”. What is the corresponding current and temperature? At which current and temperature the data of Fig. 4 were collected?
10. “The measured linewidth of the injection-locked laser is ~401.86 KHz”. How the linewidth was calculated (which line shape – Gaussian, Lorentzian or Voigt – have been assumed).
Author Response
Reviewer 1
The authors investigated the stabilization of a DFB laser by a Fabry-Perot Microcavity experimentally. Although the spectral linewidth achieved is not particular spectacular, the paper describing the experimental approach is worth to be published if it is revised properly.
(R1.1) In the Introduction, the authors should also cite some papers on the stabilization of DFB lasers by external cavities, such as Liang et al. Nature Communications DOI: 10.1038/ncomms8371, Lewoczko-Adamczyk et al OPTICS EXPRESS DOI:10.1364/OE.23.009705, and others.
Thank you for your positive comment and pointing out the relavent references, We have added them as Ref .26 and Ref. 27] in the revised manuscript. (Page 1, line 43)
(R1.2) How the theoretical linewidth of the FP cavity was estimated?
The linewidth of the FP cavity was estimated as , where n is refractive index, d is cavity length, r is surface reflectivity and c is vacuum speed of light. We use the coating reflectivity for this estimation, which agrees with the measured linewidth and shows a high fabrication quality. We have added this model in the revised manuscript. (Page 3, line 69)
(R1.3) Explain “Q factor 1.6166×104”.
We calculated the Q factor using the measured linewidth. We have corrected the excessive effective digit in the revised manuscript. (Page 4, line 73)
(R1.4) Describe the DFB laser (cavity length, type of grating, coupling coefficient, facet reflectivities etc.) Which electrical driver of the DFB laser has been used?
We are not sure of the design of the DFB laser. However, being a DFB laser, the cavity length is very short with Bragg gratings and a phase shift in the center. This laser is commercial available (model number: LSDLD155-5-S-0-2-SMFA). The model of the current driver is from wavelength electronics, inc., part number: LDTC0520. (Page 4, line 77)
(R1.5) What does it mean: “same polarization of the DFB laser”
Sorry for the confusion in the presentation. We meant the light is fed back to the DFB laser with the same polarization as its output. We have clarify this point in the revised manuscript. (Page 4, line 81)
(R1.6) “It can be seen that the output wavelength of the free-running laser significantly goes up as the temperature of laser increases (black line) due the temperature dependence of cavity length.” There are two effects affecting the laser wavelength: the temperature dependencies of the effective refractive index and the cavity length. Estimate the contributions of both effects.
Thank you for pointing out the effect of the effective refractive index, and we have included it in the revised manuscript. We think both effect contribute. (Page 5, line 108).
(R1.7) “62 mV output power level”, What is the corresponding current and temperature? At which current and temperature the data of Fig. 4 were collected?
Sorry for the confusion. It was the voltage of the photo diode, which is inaccurate and unnecessary. We have corrected in the revised manuscript. (Page 5, line 109; Page 6, line 121).
(R1.8) “The measured linewidth of the injection-locked laser is ~401.86 KHz”. How the linewidth was calculated (which line shape – Gaussian, Lorentzian or Voigt – have been assumed).
The linewidth is estimated assuming a Lorentzian line shape, and we have added this point in the revised manuscript. (Page 6, line130)
Author Response File: 
Author Response.docx
Reviewer 2 Report
Review of the manuscript „Self-Injection Locking of a Distributed Feedback Laser Diode using a high-finesse Fabry-Perot Microcavity“
The authors report about the effect of a Fabry-Perot micro-cavity (FPMC), manufactured by them, which was put into an external feedback setup of a DFB diode laser at about 1.5 µm lasing wavelength. The manuscript contains two main parts: (1) the fabrication and characterization of the FPMC and (2) the investigation of the effect of the FPMC in two diode laser setups.
The language of the paper is clear; the second part about the effect of the FPMC is clearly written and scientifically sound. The first part about the manufacturing process of the FPMC, however, needs some additional explanation, since there are several ambiguities. Therefore, I recommend publication after a revision of the manuscript.
My questions and comments regarding the manuscript are:
(1) Abstract: The authors mention that this is “the first demonstration of the injection laser locking”. I recommend to weaken this statement, because one could argue that similar ideas with slightly different technical implementations of the setup have been demonstrated already, e.g., https://0-doi-org.brum.beds.ac.uk/10.1002/lpor.201500214 , https://0-doi-org.brum.beds.ac.uk/10.1007/s00340-015-6281-z , https://0-doi-org.brum.beds.ac.uk/10.1364/OE.23.009705 .
(2) Page 1, line 27: When the authors write “in recent years [1-9]” it is conspicuous that the newest paper from [1-9] is at least 20 years old.
(3) Page 2, paragraph 2: Regarding the reflective coating of the FPMC:
- How were the layers coated, by sputtering?
- Why are the overall thicknesses different for both sides? Is the table in the supplement valid for front or rear side? Which side points to the impinging laser beam?
- The authors mention, that each layer of Ta2O5 and SiO2 is polished separately. Is this correct? How is it polished (flying cutter etc.)? Why is polishing necessary, how rough is the unpolished coating?
- The authors mentioned that the entire mirror is transferred to the LN crystal after its formation. Why is the mirror not grown/sputtered/evaporated directly onto the polished crystal? How is the mirror stack attached to the crystal? On which temporary material is the mirror grown?
- Which algorithm has been used to optimize the mirror stack?
(4) Page 3, line 70: The number format of the Q factor seem to be mixed up, please check.
(5) Page 3, line 73: “Fig. 2” instead of “Fig. 3”.
(6) Figure 2: - The term “FPMC” should be used in the scheme instead of FP Cavity to be consistent with the text.
- The left collimator is rather a telescope optics due to its task to widen the beam instead of collimating.
(7) Page 3, lines 82ff: How have the authors measured the characteristics before mode locking: Same setup without FPMC or in another setup (e.g., directly in front of the OSA)?
(8) Page 3, line 84f: Is the lower threshold of the locked laser really linked directly to the mode locking? Is it not mainly an effect of the external feedback?
(9) Page 4, line 97: “single-transverse”
(10) Page 4, line 102: “62 mV output power level” this information is not explained. Is it the output of the PD? Is there any reference to compare with?
(11) Figure 3b-d: Which current was chosen for the laser, which are compared? (How far from the threshold?)
(12) Figure 3c: All other spectra are displayed with frequency axis, as the FWHM values, too. For consistency throughout the paper please consider to add a frequency scale on the x-axis.
(13) Page 5, lines 115 and 122: The meaning of the phrase “not shown in Fig. 3” and “not shown in Fig. 2” are not clear. (data, position in setup etc.)
(14) Page 6, line 147: Is it correct, that these information are available at www.nature.com/reprints ?
Author Response
Reviewer 2
(R2.1) Review of the manuscript, “Self-Injection Locking of a Distributed Feedback Laser Diode using a high-finesse Fabry-Perot Microcavity”
The authors report about the effect of a Fabry-Perot microcavity (FPMC), manufactured by them, which was put into an external feedback setup of a DFB diode laser at about 1.5 µm lasing wavelength. The manuscript contains two main parts: (1) the fabrication and characterization of the FPMC and (2) the investigation of the effect of the FPMC in two diode laser setups.
The language of the paper is clear; the second part about the effect of the FPMC is clearly written and scientifically sound. The first part about the manufacturing process of the FPMC, however, needs some additional explanation, since there are several ambiguities. Therefore, I recommend publication after a revision of the manuscript.
My questions and comments regarding the manuscript are:
(R2.1) Abstract: The authors mention that this is “the first demonstration of the injection laser locking”. I recommend to weaken this statement, because one could argue that similar ideas with slightly different technical implementations of the setup have been demonstrated already, e.g., https://0-doi-org.brum.beds.ac.uk/10.1002/lpor.201500214 , https://0-doi-org.brum.beds.ac.uk/10.1007/s00340-015-6281-z , https://0-doi-org.brum.beds.ac.uk/10.1364/OE.23.009705 .
We truly agree with your comment, and corrected this statement in the revised manuscript. We also included the above references.
(R2.2) Page 1, line 27: When the authors write “in recent years [1-9]” it is conspicuous that the newest paper from [1-9] is at least 20 years old.
We have corrected this term. (Page 7, Line 158)
(R2.3) Page 2, paragraph 2: Regarding the reflective coating of the FPMC:- How were the layers coated, by sputtering?
The coating layers are coated by evaporation. We have added this information in the manuscript (Page 3, line 62).
- Why are the overall thicknesses different for both sides?
Sorry for the confusion. Actually the total thicknesses are the same for both sides, with thickness for the Ta2O5 and thickness for the SiO2 material, respectively.
Is the table in the supplement valid for front or rear side?
The appendix is valid for both front or rear sides.
Which side points to the impinging laser beam?
The front side point to the laser. We have added this information (Page 4, line 80).
- The authors mention, that each layer of Ta2O5 and SiO2 is polished separately. Is this correct?
Sorry for the confusion. We actually meant the lithium niobate crystal was polished before coating. We have clarified this point in the revised manuscript. (Page 2, line 56)
How is it polished (flying cutter etc.)?
Polishing is done by fine mechanical polishing. We have added this information in the revised manuscript. (Page 2, line 56).
Why is polishing necessary, how rough is the unpolished coating?
Sorry for the confusion. We did not polish the coating, but the LN facets.
- The authors mentioned that the entire mirror is transferred to the LN crystal after its formation. Why is the mirror not grown/sputtered/evaporated directly onto the polished crystal? How is the mirror stack attached to the crystal? On which temporary material is the mirror grown?
Sorry again for the confusion. We did not transfer the mirrors, but they are coated to the LN crystal facets by the evaporation method. (Page 2, line 56)
- Which algorithm has been used to optimize the mirror stack?
Our membrane is designed with TFCalc software. We have added this information in the manuscript (Page 2, line 50).
(R2.4) Page 3, line 70: The number format of the Q factor seem to be mixed up, please check.
Thank you for pointing out this typo. It has been corrected in the revised manuscript. (Page 4, line 73).
(R2.5) Page 3, line 73: “Fig. 2” instead of “Fig. 3”.
Thank you for pointing out this typo. It has been corrected in the revised manuscript. (Page 4, line 78).
(R2.6) Figure 2: - The term “FPMC” should be used in the scheme instead of FP Cavity to be consistent with the text. - The left collimator is rather a telescope optics due to its task to widen the beam instead of collimating.
Thank you for the detailed suggestions. They are corrected in the revised manuscript. (Page 4, Figure 2).
(R2.7) Page 3, lines 82: How have the authors measured the characteristics before mode locking: Same setup without FPMC or in another setup (e.g., directly in front of the OSA)?
As a reference, we routed the beam directly into the spectrometer without forming a loop. We have added this information in the manuscript. (Page 4, line 89)
(R2.8) Page 3, line 84: Is the lower threshold of the locked laser really linked directly to the mode locking? Is it not mainly an effect of the external feedback?
The reviewer is correct. The lower of the threshold is mainly due to the external feedback. We have revised the corresponding sentence (Page 4, line 90).
(R2.9) Page 4, line 97: “single-transverse”
We have revised the corresponding sentence (Page 5, line 103).
(R2.10) Page 4, line 102: “62 mV output power level” this information is not explained. Is it the output of the PD? Is there any reference to compare with?
We have revised the corresponding sentence to make the information clear (Page 5, line 109).
(R2.11) Figure 3b-d: Which current was chosen for the laser, which are compared? (How far from the threshold?)
The current is 24.8 mA. The threshold of current after injection locking is 5mA. We have added this information in the manuscript (Page 5, line 100).
(R2.12) Figure 3c: All other spectra are displayed with frequency axis, as the FWHM values, too. For consistency throughout the paper please consider to add a frequency scale on the x-axis.
For the measurement in Fig. 3c, we think the labelling the x-axis as wavelength is clearer than frequency.
(R2.13) Page 5, lines 115 and 122: The meaning of the phrase “not shown in Fig. 3” and “not shown in Fig. 2” are not clear. (data, position in setup etc.)
We have removed the corresponding sentence.
(R2.14) Page 6, line 147: Is it correct, that these information are available at www.nature.com/reprints ?
We have deleted the sentence in the manuscript.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors revised the paper according to my suggestions. However, there is still one issue which should be clarified. The authors write in the response letter:
“Thank you for pointing out the effect of the effective refractive index, and we have included it in the revised manuscript. We think both effect contribute. “
What does it mean “we think”? Did the authors check it? I think they didn’t. ). The contribution due to the temperature dependence of the effective index is more than one order of magnitude larger than the contribution to the thermal expansion! Hence the latter can be neglected!
The authors should calculate the contributions of both effects using the Bragg condition (1/Lambda=2*neff/lambda) where Lambda is the grating period, the known thermal expansion coefficient of GaAs, the temperature derivative of the effective refractive index which ranges between 2 and 4 x 10**-4 /K, and the effective index of the order 3.3 (albeit the group index should be used.
Author Response
Reviewer1
(R1.1)“Thank you for pointing out the effect of the effective refractive index, and we have included it in the revised manuscript. We think both effect contribute. “
What does it mean “we think”? Did the authors check it? I think they didn’t.). The contribution due to the temperature dependence of the effective index is more than one order of magnitude larger than the contribution to the thermal expansion! Hence the latter can be neglected!
The authors should calculate the contributions of both effects using the Bragg condition (1/Lambda=2*neff/lambda) where Lambda is the grating period, the known thermal expansion coefficient of GaAs, the temperature derivative of the effective refractive index which ranges between 2 and 4 x 10-4 /K, and the effective index of the order 3.3 (albeit the group index should be used.
We checked the contribution of both effects, and got the same result as you pointed out. That is, temperature dependence of the effective index is more than one order of magnitude larger than the contribution to the thermal expansion. Thank you for this good point, and we have revised the manuscript accordingly. (Page 4, line 92-93)
Author Response File: 
Author Response.pdf
