See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/24433694

Stable, 96 W, continuous-wave, single-frequency, fiber-based green source at

532 nm

Article in Optics Letters · June 2009

DOI: 10.1364/OL.34.001552 · Source: PubMed

CITATIONS READS
70 138

3 authors:

Goutam K Samanta Suddapalli Chaitanya Kumar

Physical Research Laboratory ICFO Institute of Photonic Sciences
176 PUBLICATIONS 1,864 CITATIONS 193 PUBLICATIONS 2,298 CITATIONS

SEE PROFILE SEE PROFILE

Majid Ebrahim-Zadeh
ICFO Institute of Photonic Sciences
474 PUBLICATIONS 6,623 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Nonlinear frequency conversion View project

Mid-TECH ITN project View project

All content following this page was uploaded by Suddapalli Chaitanya Kumar on 01 June 2014.

The user has requested enhancement of the downloaded file.

 May 15, 2009 / Vol. 34, No. 10 / OPTICS LETTERS 1561

Stable, 9.6 W, continuous-wave, single-frequency,

fiber-based green source at 532 nm
G. K. Samanta,1,* S. Chaitanya Kumar,1 and M. Ebrahim-Zadeh1,2
1
ICFO-Institut de Ciencies Fotoniques, Mediterranean Technology Park, 08860 Castelldefels, Barcelona, Spain
2
Institucio Catalana de Recerca i Estudis Avancats (ICREA), Passeig Lluis Companys 23, Barcelona 08010, Spain
*Corresponding author: [email protected]
Received January 7, 2009; revised April 3, 2009; accepted April 6, 2009;
posted April 20, 2009 (Doc. ID 106134); published May 12, 2009
We present a stable, high-power, cw, single-frequency green source in a compact and practical design based
on a simple single-pass second-harmonic generation of a cw ytterbium fiber laser at 1064 nm in MgO-doped
periodically poled stoichiometric LiTaO3. Using a 30-mm-long crystal containing a single grating, we have
generated 9.64 W of cw radiation at 532 nm with a fundamental power of 29.5 W at a single-pass conversion
efficiency of 32.7%. The output power is naturally stable with a peak-to-peak fluctuation of 7.6% over the
first 8 h and 9% over 13 h. Over the entire range of fundamental powers, the generated green output is
single frequency with an instantaneous linewidth of 6.5 MHz and frequency stability of ⬍32 MHz over 30
min and has a TEM00 spatial profile with M2 ⬍ 1.33. © 2009 Optical Society of America
OCIS codes: 140.7300, 140.3510, 190.2620, 190.4400.

High-power, cw, and single-frequency green sources cw, green sources based on cw fiber laser technology,
are of great interest for a variety of applications, in- offering stable, high-power, and single-frequency out-
cluding pumping of Ti:sapphire lasers and cw singly put with high spatial beam quality, mandatory re-
resonant optical parametric oscillators (SROs) quirements for the above applications, has not been
pumped in the visible [1]. To date, practical realiza- extensively explored. In this Letter, we describe such
tion of such sources has relied almost exclusively on a source using a simple design based on an external
internal second-harmonic generation (SHG) of cw single-pass SHG of a cw, single-frequency, ytterbium
Nd-based solid-state lasers. While these techniques (Yb) fiber laser at 1064 nm in MgO:sPPLT. The
have been proven highly effective, the attainment of source can deliver 9.64 W of cw single-frequency out-
stable, high-power, and single-frequency operation put power at 532 nm at a conversion efficiency as
generally requires elaborate system designs involv- high as 32.7% in a TEM00 spatial profile with peak-
ing intricate cavity configurations for internal dou- to-peak power stability of 9% over 13 h and frequency
bling, careful management of thermal effects, and ac- stability of ⬍32 MHz over 30 min. Together with its
tive stabilization, resulting in relatively high high passive stability and single-frequency perfor-
complexity and cost. It would be desirable to devise mance, the device offers an attractive, compact, and
alternative approaches for the development of such cost-effective source for many applications, including
sources using more simplified and cost-effective tech- pumping of Ti:sapphire lasers and high-power cw
niques. An attractive approach is the external SHG of SROs.
high-power cw IR lasers in quasi-phase-matched The configuration of the experimental setup is
ferroelectric materials such as MgO-doped periodi- similar to our previous work [5]. The fundamental
cally poled LiNbO3 (MgO:PPLN) [2], MgO-doped pe- pump source is a 30 W, cw, single-frequency Yb fiber
riodically poled stoichiometric LiTaO3 (MgO:sPPLT) laser (IPG Photonics, YLR-30-1064-LP-SF) at 1064
[3,4], and periodically poled KTiOPO4 (PPKTP) [5]. nm. The laser delivers a linearly polarized output
The onset of photorefractive damage under expo- beam with a diameter of 3 mm, M2 factor of ⬍1.01,
sure to visible radiation has limited the use of and a nominal linewidth of 89 kHz. Using a waveme-
MgO:PPLN mainly to the near- and mid-IR. On the ter (HighFinesse WS-U 30), we measure the fre-
other hand, MgO:sPPLT [3,4] and PPKTP [5] are at- quency stability of the laser to be ⬍120 MHz over 1 h
tractive for frequency conversion in the visible be- and ⬍50 MHz over 30 min. The nonlinear crystal is
cause of the increased resistance to photorefractive 30 mm MgO:sPPLT (HC Photonics, Taiwan), contain-
damage and relatively high effective nonlinearities ing a single grating 共⌳ = 7.97 ␮m兲, and housed in an
共deff ⬃ 10 pm/ V兲. In PPKTP, thermal issues arising oven with a temperature stability of ±0.1° C. The
from the absorption of green radiation together with crystal faces have low reflectivity 共R ⬍ 0.5%兲 at 1064
gray tracking have so far limited cw green generation nm and high transmission 共T ⬎ 99%兲 at 532 nm [1]. A
to 6.2 W [5]. MgO:sPPLT has demonstrated to be the dichroic mirror (R ⬎ 99% at 1064 nm; T ⬎ 94% at 532
most promising material owing to its high photore- nm) separates the generated green from the input
fractive damage threshold [6] and large thermal con- fundamental.
ductivity [7] to handle high optical powers. Earlier To optimize the SHG, we used several focusing con-
reports of green generation in MgO:sPPLT include ditions corresponding to different values of the focus-
single-pass SHG of a 91.5 W cw Nd:YAG laser, pro- ing parameter, ␰ = L / b [8]. Here L is the crystal
2
viding 16.1 W at 532 nm at 17.6% efficiency [4]. How- length and b = kwop is the pump confocal parameter,
ever, to our knowledge, the development of compact, with k = 2␲np / ␭p, where np, ␭p, and wop are the refrac-
0146-9592/09/101561-3/$15.00 © 2009 Optical Society of America
1562 OPTICS LETTERS / Vol. 34, No. 10 / May 15, 2009

tive index, wavelength, and waist radius of the fun- power and the corresponding PM temperature. On
damental beam inside the crystal, respectively. We the other hand, given the significant linear absorp-
measured the maximum generated second-harmonic tion of 0.17%/cm at 1064 nm, the role of fundamental
(SH) powers and corresponding optimum phase- absorption to crystal heating cannot be entirely ig-
matching (PM) temperatures using seven different nored. To verify this contribution, under the focusing
focusing conditions, ␰ = 0.32, 0.81, 1.23, 1.74, 2.48, condition of ␰ = 1.74, we rotated the fundamental po-
4.50, and 6.60 at a fixed fundamental power of 29.5 larization to generate a green power of less than 100
W at the input to the crystal. The results are shown mW at 29.5 W of fundamental. By increasing the fun-
in Fig. 1. For weak focusing 共␰ ⬍ 2.48兲, the maximum damental power from 1 to 29.5 W, we observed a drop
SH power increases with higher ␰, whereas for tight in the PM temperature from 52.5° C to 51.6° C,
focusing 共␰ ⬎ 2.48兲, it decreases with increasing ␰. In- clearly indicating still a significant contribution of
terestingly, the extrapolated power curve has a clear fundamental absorption to thermal effects. These
peak near ␰ ⬃ 2.84, corresponding to the theoretical studies thus confirm that thermal effects in the
prediction for optimum SHG in the cw (or long-pulse) MgO:sPPLT crystal were owing neither to GRIIRA
limit [8]. We obtained a maximum SH power of nor TPA but a result of intrinsic linear absorption of
9.64 W at ␰ = 2.48 共wop ⬃ 31 ␮m兲, corresponding to a IR and green. It has been suggested that the maxi-
single-pass conversion efficiency of 32.7%. mum available SH power is limited by the thermal
It is also clear in Fig. 1 that the PM temperature effects resulting from either only the absorption of
decreases with tighter focusing. This is to be ex- fundamental [9] or only the SH power [4]. Our
pected since crystal heating effects owing to various present studies together with our earlier report [10]
absorption mechanisms including green-induced IR confirm that the thermal effects are owing to the ab-
absorption (GRIIRA) of fundamental, two-photon ab- sorption of both fundamental and SH power, with the
sorption (TPA) of fundamental and green, and linear major contribution from the green.
absorption at both wavelengths are stronger under From the measurement of green power in Fig. 1,
tight focusing, leading to a greater thermal load in the normalized conversion efficiency was calculated
the crystal and therefore necessitating lower exter- to vary from 0.42%/W at ␰ = 0.32 to 1.26%/W at ␰
nally applied heat to the sample. GRIIRA is not ex- = 6.60, with a maximum of 1.70%/W at ␰ = 2.48. The
pected to make a significant contribution to crystal normalized efficiency is not limited by fundamental
heating at these power densities because of its sup- linewidth, since the spectral acceptance of the 30 mm
pression owing to MgO doping [6]. To investigate the crystal calculated from Sellmeier equations [11] is
role of TPA, we recorded crystal transmission at fun- 0.082 nm (FWHM), far wider than the fiber laser
damental power levels up to 30 W and in the green at linewidth of 89 kHz.
up to 9 W using a commercial cw green source (Co- To characterize the power scalability of the SHG,
herent, Verdi-10). We obtained a linear increase in we recorded the SH power and efficiency at ␰ = 2.48
the transmitted power with input power at both 共wop ⬃ 31 ␮m兲 up to the maximum available fiber la-
wavelengths, thus also confirming the absence of ser power, as shown in Fig. 2. We obtained 9.64 W of
TPA. From the measurements, we deduced a linear green for the full fundamental power of 29.5 W at a
absorption loss of 1.58%/cm at 532 nm and 0.17%/cm single-pass efficiency of 32.7%. The quadratic in-
at 1064 nm, implying a far higher contribution to crease in the SH power and the corresponding linear
crystal heating from the absorption of green than variation in efficiency are maintained up to a funda-
fundamental. This is also evident in Fig. 1, where at mental power of 22 W (Fig. 2), after which saturation
a fixed fundamental power of 29.5 W, we observe a occurs. The saturation effect is also evident from the
strong correlation between the generated green

Fig. 1. Maximum SH power and corresponding PM tem-

perature as a function of the focusing parameter, L / b. The Fig. 2. Dependence of the measured cw SH power and the
vertical dashed line corresponds to the optimal focusing corresponding conversion efficiency on the incident funda-
condition 共L / b ⬃ 2.84兲 prescribed by Boyd and Kleinman mental power. Inset, variation in the SH power as a func-
[8]. tion of square of the fundamental power.
 May 15, 2009 / Vol. 34, No. 10 / OPTICS LETTERS 1563

deviation of the linearity of the SHG power with the

square of fundamental power (inset of Fig. 2) and is
attributed to the pump depletion, backconversion,
and thermal phase-mismatch effects in the
MgO:sPPLT crystal. We believe that further in-
creases in the SH power beyond 9.64 W will be pos-
sible using a pump beam waist of ⬃29 ␮m 共␰ = 2.84兲
and improved thermal management to overcome
saturation.
The frequency stability of the generated green,
measured at 9.64 W using a wavemeter (HighFi-
nesse, WS-U 30), is shown in Fig. 3. Under free-
running conditions and without thermal isolation,
the green power exhibits a peak–peak frequency fluc-
tuation of ⬍115 MHz over 90 min, with a short-term
stability of ⬍32 MHz over 30 min. The transmission
fringes obtained through a confocal scanning inter-
ferometer (FSR= 1 GHz, finesse= 400), shown in the
inset of Fig. 3, also verify the single-frequency spec- Fig. 4. (Color online) Green output power stability at
trum with an instantaneous linewidth of ⬃6.5 MHz. 9.64 W over 13 h and (inset) far-field TEM00 energy distri-
Similar behavior was observed throughout the pump- bution and intensity profiles of the generated green beam.
ing range with the same linewidth, confirming robust In conclusion, we have generated 9.64 W of cw
single-mode operation at all fundamental powers. single-frequency radiation at 532 nm with a conver-
The power stability near the maximum green sion efficiency as high as 32.7% using a simple single-
power of 9.64 W is shown in Fig. 4, demonstrating a pass SHG of a cw Yb fiber laser in MgO:sPPLT near
peak–peak fluctuation of 7.6% over the first 8 h and room temperature. The source exhibits good long-
9% over 13 h. The power fluctuation is attributed term passive frequency and power stability and a
mainly to changes in the laboratory environment, so TEM00 spatial profile with M2 ⬍ 1.33. With more
further improvements in power stability, below 3%, stringent thermal control and optimized focusing,
are expected through thermal isolation of the system further improvements in green power and efficiency
and better temperature control. The far-field energy as well as power scaling are feasible.
distribution of the green beam at 9.64 W, together
with the intensity profile and the Gaussian fits along We acknowledge support from the Spanish Minis-
the two orthogonal axes, are shown in the inset of try of Education and Science through grant
Fig. 4. Using a focusing lens 共f = 25 cm兲 and scanning TEC2006-12360 and the Consolider project
beam profiler, we measured M2 values of the green (CSD2007-00013).
beam to be M2x ⬃ 1.29 and M2y ⬃ 1.23 with an ellipticity References
of ⬃0.96, confirming the TEM00 spatial mode. Similar
1. G. K. Samanta, G. R. Fayaz, and M. Ebrahim-Zadeh,
M2 values were measured at different input power Opt. Lett. 32, 2623 (2007).
levels, showing a small variation in M2x from 1.11 to 2. H. Furuya, A. Morikawa, K. Mizuuchi, and K.
1.29 and M2y from 1.17 to 1.33. Yamamoto, Jpn. J. Appl. Phys. Part 1 45, 6704 (2006).
3. S. V. Tovstonog, S. Kurimura, and K. Kitamura, Appl.
Phys. Lett. 90, 051115 (2007).
4. S. V. Tovstonog, S. Kurimura, I. Suzuki, K. Takeno, S.
Moriwaki, N. Ohmae, N. Mio, and T. Katagai, Opt.
Express 16, 11294 (2008).
5. G. K. Samanta, S. Chaitanya Kumar, M. Mathew, C.
Canalias, V. Pasiskevicius, F. Laurell, and M. Ebrahim-
Zadeh, Opt. Lett. 33, 2955 (2008).
6. K. Kitamura, Y. Furukawa, S. Takekawa, T. Hatanaka,
H. Ito, and V. Gopalan, Ferroelectrics 257, 235 (2001).
7. N. E. Yu, S. Kurimura, Y. Nomura, and K. Kitamura,
Jpn. J. Appl. Phys. Part 2 43, L1265 (2004).
8. G. D. Boyd and D. A. Kleinman, J. Appl. Phys. 39, 3597
(1968).
9. D. S. Hum, R. K. Route, G. D. Miller, V. Kondilenko, A.
Alexandrovski, J. Huang, K. Urbanek, R. L. Byer, and
M. M. Fejer, J. Appl. Phys. 101, 093108 (2007).
10. G. K. Samanta and M. Ebrahim-Zadeh, Opt. Express
16, 6883 (2008).
Fig. 3. Frequency stability of green beam at 9.64 W over 11. A. Bruner, D. Eger, M. B. Oron, P. Blau, M. Katz, and
90 min and (inset) corresponding single-mode spectrum. S. Ruschin, Opt. Lett. 28, 194 (2003).

View publication stats