An NIR laser frequency comb for

Document Sample
An NIR laser frequency comb for Powered By Docstoc
					An NIR laser frequency comb for high precision Doppler planet surveys
Center for Astrophysics and Space Astronomy University of Colorado

Steve Osterman

Scott Diddams (NIST) Frank Quinlan (NIST) John Bally (CU) Jian Ge (UF)

+
Steve Osterman NIR laser frequency comb

UF
BDEP 20 July 2009

Overview
Why look for planets around low mass stars, and why look in the NIR? Why do we need Laser Frequency Combs The CU/NIST/UF LFC Program Where can we go with this?

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Why would we look for planets around low mass stars, and why in the NIR?
Larger RV signature for a given planet mass in the habitable zone Large number of host stars within 10pc Cool stars brighter in the NIR No shortage of narrow spectral features
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Increased RV signature:
Low Temperature, Low Mass Host: Habitable zone is closer to the host, increasing RV signature Lower host mass increases RV signature Tighter orbit leads to shorter period (weeks)
Stellar RV for earth mass planet in the habitable zone. Derived from Kasting (1993, fig. 15).

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Large number of host stars within 10pc
250 200

150

100

50

Data from most recent RECONS survey values (Jan 2009) showing predominance of class M stars within 10 pc.

0

WD O Steve Osterman

B

A

F

G

K

M

L

T

P BDEP 20 July 2009

NIR laser frequency comb

Cool, low mass stars are brighter in the NIR
M6 Model Spectrum

Optical RV surveys are limited to stars more massive than early M dwarfs (>0.3 Msun); lower mass stars are too faint in visible light for optical RV surveys

(courtesy of Peng Jiang, 2009)

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Distinct, Narrow Features
M and L dwarfs have numerous sharp absorption features in the H and J bands

Fe absorption features in J and H band, from Cushing, 2003 Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Precision NIR Radial Velocity Spectroscopy would allow us to address
How common are planets around low mass stars (K,M)? What are the masses and orbital parameters of these planets? How do the total planetary mass and orbital parameters depend on stellar mass Origins and early evolution of planetary systems Physical processes and initial conditions that produce different types of systems And finally Where are the potentially habitable planets

The only real issue here is using the word precision when discussing NIR spectroscopy…
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Current wavelength standards for the NIR
Historically, ‘precision’ spectroscopy in the NIR has been any thing but precise Visible precision spectrographs have a variety of tools available – most notably I2 absorption cells and ThAr emission line lamps. Reported RV precision as low as 60cm/sec NIR (1-2 m) has typically been limited to > 80m/s, with relatively poor performance from Absorption cells Emission line lamps Telluric lines
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Current wavelength standards for the NIR
Absorption cells Complicate the spectra – especially shallow absorption features. Attenuate science signal. Overlapping lines lead to unresolved features. Most current cells limited to visible ranges
Fe II 1608 and 1611 lines from Keck/ HIRES showing spectrum with and without I2 cell. (From Griest, 2009) Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Current wavelength standards for the NIR
Absorption cells – molecular sources (C2H2, 12CO, 13CO and HCN) provide limited coverage. Cascaded cells possible but… Limited coverage (1.51-1.63 m requires 4 species). Complicate the spectra. Attenuate science signal. Overlapping lines lead to unresolved features.
Mahadevan and Ge, ApJ 692:1590–1596, 2009

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Current wavelength standards for the NIR
Emission line lamps ThAr lamps dim with limited lines in H band Multi lamp systems (e.g. PRVS proposed a mix of Th-Ar, Ar, Kr, Ne, and Xe penlamps) Telluric lines Few or none in H- band (that’s why we work there in the first place Limited to >25m/s beyond 2 m or below 1.1 m While all these methods provide coverage over limited bands, these are as or more complex than ThAr or I2 in the visible without the heritage
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

NIR Laser Frequecncy Comb for High Precision Spectroscopy
An ideal wavelength calibration source would provide…
A high density array of uniformly spaced, uniformly bright lines Lines can be traced to the standard second. The precision and long term stability of this source should exceed the ultimate precision of the spectrograph.

The output frequencies emitted from a mode-locked laser create a high precision “optical frequency ruler.”

fn = nfr + f0
fr is the repetition rate of the femtosecond laser and both fr and f0 are controlled by a high stability atomic source. If fn and f0 are tied to a compact rubidium atomic reference accuracy of 10-11 can be achieved.
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Frequency combs span the visible and NIR
LFC spectral envelopes :
(a) Broadband Ti:sapph

(1-5 GHz)
(b) Yb:KYW plus nonlinear

microstructure fiber (250MHz)
(c) narrow band 800 nm

Er:fiber plus nonlinear microstructure fiber. (250MHz) Recent spectra generated with Er:fiber and nonlinear fiber extend from 1100 nm to beyond 2200 nm.
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

1.55

m Er:Fiber+filter cavity setup
Filter cavity servo PZT
Optical circulator (isolator)

10 MHz reference for locking comb spacing and offset frequency

Filter cavity

Filtered comb output in single mode fiber

Original 250 MHz comb & filter cavity transmission

Filtered 12.5 GHz comb (0.1nm, or /

=15,500 at 1.55μm)

Filter cavity selects one mode of every 50 to generate 12.5 GHz-spaced comb.
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Filter cavity performance
Optical spectrum

High res. close-in

Loss per mode at center of spectrum ~4 dB. Greater loss in the wings is due to filter cavity dispersion.

Mode suppression characterization
Experimental setup
Filter cavity servo

Optical circulator (isolator)

PZT

TBPF

EDFA PC Spectrum Analyzer

Filter cavity

Tunable laser source Tunable laser is tuned from 1525 nm to 1590 nm to measure side mode suppression.
PC: polarization controller EDFA: Erbium-doped fiber amplifier TBPF: tunable bandpass filter PZT: piezoelectric transducer

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Spur suppression measurements
Optical domain
Tunable laser source (TLS) Comb modes

Photodetected beat between filtered comb and TLS
beat 1 beat 2

Resonant higher order spatial mode Steve Osterman

“blue” nearest neighbor

“red” nearest neighbor

TLS wavelength: 1560 nm

NIR laser frequency comb

BDEP 20 July 2009

Spur suppression measurements
Nearest neighbor (NN) suppression Higher order mode (HOM) suppression

Reduction in coating reflectivity
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Current NIST Configuration
second filter cavity increases spur suppression
Filter cavity servo Optical circulator (isolator) PZT

Filter cavity

PC Filter cavity C- and L-band amplification

Current performance:
>34dB suppression single cavity, m=50 filter ratio (12.5GHz/250MHz) 70nm single cavity coverage (1530-1600nm, defined as >10% of maximum transmission) Second cavity in series eliminates HOSM, should improve NN spression, with comparable coverage
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

FIRST + Frequency Comb

Fiber fed, 55,000 / SIG Spectrograph, 1.2-2.4 m Dual calibration feeds (cal. + ref.)
Steve Osterman

Comb ready for off site testing in October Anticipate operations at APO in early 2010
BDEP 20 July 2009

NIR laser frequency comb

FIRST + Frequency Comb: Single order and narrow band simulations

M6 model spectrum (black) and LFC (red, offset and stretched)

Fiber fed, 55,000 / SIG Spectrograph, 1.3-1.8 m Dual calibration feeds (cal. + ref.)
Steve Osterman

Comb ready for off site testing in October Anticipate operations at APO in early 2010
BDEP 20 July 2009

NIR laser frequency comb

Suppression modeling

For a 250MHz comb feeding a 55,000 resolution spectrograph, ~12 comb lines fall within 1FWH The registered line center will be influenced by the following
• • •

Detuning: Nearest and next nearest neighbor asymmetry Detuning: Comb profile deformation (Braje, 2008) Higher order Spatial modes – both detuning effects, as well as any uncontrolled effects
NIR laser frequency comb BDEP 20 July 2009

Steve Osterman

Toleranceing
Comb to filter misalignment corresponds to apparent line shift:
• •

Filter cavity passes different ratios of left and right nearest neighbors Filter cavity reshapes central line (Braje, 2008)

Increased comb natural line width increases sensitivity to small offsets

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Toleranceing
Comb to filter misalignment also modulates transmitted power This provides a robust measure of offset for relatively narrow lines

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Toleranceing
Comb to filter misalignment also modulates transmitted power This provides a robust measure of offset for relatively narrow lines

Current line width is ~ 0.1MHz at 1.55 m, 55,000 resolution 1.0MHz supports ~10cm/s
Steve Osterman

comb currently supports 1cm/s

NIR laser frequency comb

BDEP 20 July 2009

Resulting RV Sensitivity
Using an ideal spectrograph, the NIST comb can support a terrestrial planet search out to class G stars With a current spectrograph we can support a terrestrial planet search of ~5Me planets around M stars
Projected FIRST Performance Stellar RV for earth mass planet in the habitable zone. Derived from Kasting (1993, fig. 15).

Ideal Spectrograph: 1cm/s

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Testing Planet Migration Theories
Gas Giants around Solar type form only at r > 2 - 4 AU due to shear Must happen before H2 is lost to UV photo-ablation

Gravitational instability t ~ 0 - 3 Myr
(Requires high surface density disk)

Core accretion

(Requires 5 - 10 Earth mass rock/ice core)

t ~ 2 - 5 Myr

Migration follows formation In disk migration models, migration occurs in obscured, embedded phase. What is the youngest star orbited by a “hot Jupiter”? By seeing through the dust obscuring young stars, we could constrain time & mechanism of migration
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

Testing Planet Migration Theories
HH 46/47: a young embedded star at visible and IR wavelengths

NTT [O ] [S ] = 0.38, 0.65, 0.67 μm Bally & Reipurth (06 “Birth of Stars & Planets” CUP = BR06)
Steve Osterman

Spitzer H2 PAH 3.6, 4.5, 8 μm (Noriega-Crespo 04; BR06)
BDEP 20 July 2009

NIR laser frequency comb

Other projects enabled by high precision NIR spectroscopy
How common are terrestrial mass planets around low mass stars, and how many reside in the habitable zone? How and when do gas giant orbits evolve? How common are gas giant planets around post-main sequence red giant? Are “Hot Jupiters” Cannibalized by Red Giants? (IRC 10216, R Cor Bor, …) How Common are Gas Giant, Brown Dwarfs, and Red Dwarfs Around Massive Super-giants? (Aldebaran, Antares, Betelgeuse, VY Canis Majoris, …) Planetary atmospheres Stellar rotation and astroseismology Low mass spectroscopic binaries

This is not just about finding planets around M stars: By improving RV precision by 2 orders of magnitude we open up an enormous discovery space.
Steve Osterman NIR laser frequency comb BDEP 20 July 2009

The End

What next? Combs in space?
High resolution astronomical spectroscopy would benefit from space-based platform:
No atmospheric absorption and blurring No wind loading Increased thermal stability Increased pointing stability

Frequency comb technology has progressed to the point where deployment in space appears feasible: SWAP of flight qualified Er and Yb-based combs could be as low as 10 liters, 10 kg , 10 W

J H

K

L

M

N

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Reduced stellar noise
M dwarfs may show less jitter than more massive stars M dwarf activity probably limited to only the youngest stars
from Keck Sample, Wright (2005)

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

IR Doppler
Lunine explored the probability of discovery as a function of the radiant equivalent radius and found a distinct advantage for IR Doppler as a tool for probing the habitable zone of nearby stars

From Lunine, 2009

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Exo-planets around nearby mature K, M dwarfs
Most common type of star in the Solar vicinity Most K, M stars are single K and M spectral types:
M ~ 0.08 to ~ 0.7 Solar masses L ~ 10-4 to 0.16 Solar luminosities
Stellar Mass (M0) 0.10 0.21 0.47 0.65 0.78
Steve Osterman

Planet Mass (me) 1.0 1.0 1.0 1.0 2.0

Lum. (L0) 8e-4 7.9e-3 6.3e-2 1.6e-1 4.0e-1

Type

RHAB. (AU) 0.028 0.089 0.25 0.40 0.63

RV Period (cm/s) (days) 168 65 26 18 25 6 21 67 115 209
BDEP 20 July 2009

M8 M5 M0 K5 K0

NIR laser frequency comb

Different spectral ranges require different technologies:

Steve Osterman

NIR laser frequency comb

BDEP 20 July 2009

Ti:sapphire compared to Er:fiber
CW pump
M3 PUMP M1 M2 OC

Gain (Er-doped fiber) singlemode fiber polarizer

Isolator polarizer controller Ti:Sapphire 800 nm up to 3 GHz >500 mW 5-8 W @ 532 Required Very low <30 fs ~700 W

Center wavelength: Repetition Rate: Power out: Pump: Alignment: Noise: Pulse width: Electrical Power:

Er:fiber laser 1560 nm ~ 250 MHz 3-30 mW 100 mW @ 1480/980 nm Easy Higher, but rapidly improving 90-200 fs ~10 W

See also work of: Schnatz & Telle (PTB), Holzwarth (MenloSystems), Steve OstermanTauser (Toptica), Hartl (IMRA), Leitenstorfer (Konstanz) NIR laser frequency comb BDEP 20 July 2009

Backup slide: 10 GHz femtosecond Ti:sapphire ring laser
M1-M3 with -40 fs2 GDD Total cavity GDD -35 fs2 Total cavity length 30 mm 6.5 W 532 nm pump power 1% OC 2% OC 8 mm 0.65 W output 1.06 W output

Laser self starts

Steve Osterman

A. Bartels, NIR laser frequency combUniv. Konstanz and GigaOptics 2009 BDEP 20 July


				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:114
posted:12/24/2009
language:English
pages:38