3D Printing an External Cavity Diode Laser Housing
aa r X i v : . [ phy s i c s . e d - ph ] S e p
3D Printing an External Cavity Diode Laser Housing
E. Brekke ∗ and T. Bennett Department of Physics, St. Norbert College, De Pere, WI 54115
H. Rook
Department of Physics, Carleton College, Northfield, MN 55057
E.L. Hazlett
Department of Physics, St. Olaf College, Northfield, MN 55057 (Dated: September 24, 2020)
Abstract
The ability to control the frequency of an external-cavity diode laser (ECDL) is an essentialcomponent for undergraduate laboratories and atomic physics research. Typically the housing forthe ECDL’s diffraction grating and piezoelectric transducer is either purchased commercially ormachined from metal. Here, we present an alternative to these commonly used options that utilizes3D printing, a tool available in many physics departments. We characterize the performance of ourECDL system using atomic spectroscopy and self-heterodyne interferometry and show that it issufficient for use in undergraduate spectroscopy experiments and a number of research applicationswhere extremely narrow laser linewidths are not necessary. The performance and affordability of3D-printed designs make them an appealing option for future use. . INTRODUCTION Laser diodes are now commonplace in both undergraduate experiments and atomic re-search. While the diodes themselves are cheap and easy to obtain, stable and controllableoperation of these diodes requires implementing of an optical feedback method. Most com-monly, this needed feedback is accomplished through the use of a diffraction grating in theLittrow configuration. In this method, the diffraction grating forms an external cavity ofthe laser and, by adjusting the angle of the grating, the frequency of the resonant lightin this external cavity is controlled. By integrating a piezoelectric transducer in this sys-tem that provides feedback for the adjustment of the grating angle, a frequency-stabilizeddiode is attained. The laser housing for such an extended-cavity diode laser (ECDL) is anassembly that supports and allows for the adjustment of the essential components. Mostcommonly, this assembly is purchased commercially or is constructed from machined alu-minum components.
The latter approach works well, but requires access to machiningtools and expertise, which are often barriers for implementation.The method presented here breaks down these barriers by using 3D printing technology.With recent significant decreases in the cost of 3D printers and materials as well as access tocomputer-aided design (CAD) programs, 3D printing is now common in most educationaland industrial environments. Hence, these fabrication technologies have found a number ofuses in physics instruction and research.
In this paper, we will show that modern 3D printing techniques provide a viable alter-native for the creation of necessary components for an ECDL system. Here, we present alaser assembly that enables optical feedback with the Littrow method. As shown in Fig.1, our assembly is constructed from 3D printed materials and simple low-cost commercialcomponents. We will show that this design has a large mode-hop-free tuning range, a suffi-cient short-term stability, and the adaptability necessary to make it an excellent option forundergraduate laboratories and atomic research.
II. DESIGN
The essential design elements of an ECDL assembly provide support for a diffractiongrating, ability to adjust the vertical and horizontal location of the optical feedback beam,2nd positioning of a piezoelectric transducer to enable scanning of the laser frequency. Thegoal of the design presented here is to incorporate these elements into a 3D printed ver-sion, which requires no additional machining. For our system, the components that cannotbe 3D printed are easily available, ensuring the broad accessibility of the design. For acomplete materials list, CAD files, and detailed assembly instructions see the supplemen-tal information. We used a TazBot Lulz 6 printer, which is a Fused Filament Fabricationprinter, with PLA (Polylactic Acid) filament. Figure 1 depicts the principle of optical feedback for an ECDL and the key componentsof the design. The first-order diffraction ( m = −
1) of the emitted laser light provides thefeedback that controls the frequency of the laser operation at a desired frequency. Light ofthe same order, but of higher (lower) frequency are deflected above (below) the resonant m = − m = 0 order light is then reflected off a mirrorthat rotates with the grating to maintain alignment of the light. These optics are mountedon a rotatable base to control the frequency in a coarse manner as indicated by the color ofthe arrows. A piezoelectric transducer provides fine adjustment of the grating for electroniccontrol of the optical feedback. FIG. 1. a) The basic operation of a Littrow ECDL showing optical feedback to the laser diodefrom the m = − In our assembly, the diffraction grating is attached to the 3D housing base via super-glue3r epoxy. An 1800 lines/mm grating was used, giving a 45-degree angle of operation at 780nm. The coarse angle adjustment is done with a fine-thread set screw and spring, which canbe locked into place with two 4-40 set screws. A reflection mirror is used to ensure constantoutput steering. It is important to choose a mirror with a thin profile (less than 3 mm) inorder to prevent multiple reflections off of the grating. If the experiment is not sensitive tooutput steering, or if the laser will not be tuned extensively, the mirror can be eliminatedcompletely. This modification can be done by adjusting the CAD file or by removing thearm in post-print.The rotation of the feedback housing has a large travel range, but not all of this range isusable due to multiple reflections from the steering mirror and the limited distance of travelof the coarse adjustment screw. Both of these issues can be easily modified in the CADfiles to tailor to the base for a specific diode. For the version shown here, the coarse anglescan range is limited to 42-48 degrees with respect to the normal of the grating. Takingthese angles into account, the estimated compatible wavelength ranges are from to 743–825nm and 1115–1238 nm range with 1800 lines/mm and 1200 lines/mm grating, respectively.The edges of these ranges are dependent on the grating being mounted flush against thebase. The angle of the grating mount surface can be adjusted to accommodate wavelengthof other diodes. We have implemented this change successfully to tune a laser near 1010 nm.Elimination of the steering mirror expands the tuning range, at a sacrifice to the stabilityof the output-beam direction.Once the coarse angle alignment is set, the vertical angle is adjusted with an additionalfine-thread screw. Here the natural elasticity of the 3D printed housing is used to maintainthe alignment instead of a spring. A 5-degree vertical pitch is built into the grating mountingto ensure that proper vertical alignment occurs in the middle of an adjustable range.A piezo transducer is used for the fine angle adjustment, counterbalanced with a builtin elastic cantilever, similar to the vertical alignment system. For ABS (Acronylite Buta-diene Styrene) and PLA, the surface hardness is large enough that the piezo can tune thewavelength over several GHz with minimal hysteresis.Due to shrinkage and different slice settings, the housing for the piezo transducer maybe larger than the piezo itself. This problem can be alleviated by either printing test piecesor through the insertion of small metal shims that are easily available at any hardwarestore or online. It is important to note that although the location of the coarse horizontal4djustment assembly is not positioned for optimal scanning range, the piezo scan pivotprovides synchronous cavity mode and feedback wavelength scanning. There are various ways to secure the laser to an optical table. The base design includedin the supplemental material is shown in Fig. 1(b). This design includes a set socket fora 8-32 set screw that is compatible with an optical post or pedestal, or table clamps canbe used to secure the base to an optical table or breadboard. The base can also be easilymodified to allow for a direct mounting to a 1-inch grid optical table, as shown in Fig. 1(c).It is expected that the exact performance of the laser will depend on the material used andthe filling geometry of the print. Here PLA was shown to be effective, with the tetrahedralinfill seeming to provide extra stability. Fills between 20% and 100% were used, with thebalance of flexibility and stability being especially important for the vertical adjustment.The print parameters could be further investigated to determine the ideal characteristics forparticular applications.
III. MATERIALS AND COST
A great benefit of 3D printing is the reduction of the cost barrier to scientific investigationssuch as the Foldscope and hand-powered blood centrifuge. In this spirit, one of the goalsof our design was to keep the cost as low as possible. We estimate a cost of less than $20for 3D printing the necessary components.Table I lists the components needed for the full assembly. These components comprisethe main optics and optomechanics for the ECDL and put the cost of this project at lessthan $600, which is an order of magnitude less than the cost of a commercial ECDL. To fullyutilize the system, stabilization of the laser-diode temperature and drive current is required.In our setup, we used a commercial diode laser current driver (Thorlabs LDC202C)and temperature controller (Thorlabs TED200C) costing approximately $1000 each. Thissignificantly increases the cost of the laser system. By integrating our ECDL with home-built laser diode current drivers, piezo drivers, and temperature controllers, the costfor a complete atomic spectroscopy can be significantly reduced. One can also build a laserdiode mount with integrated thermoelectric cooling ability to replace the Thorlabs LDM21mount and its cooling ability. This mount would require either the laser diode to have thesame pointing as the LMD21 mount or for the CAD files to be adjusted.5his setup will not replace the commercial ECDL in all research applications, but it canreplace it in simple spectroscopy experiments and experimental components. In addition,this design is extremely beneficial for upper division lab courses, giving students access tolaser and spectroscopy experiments. The cost advantages and ease of use make this systeman excellent option for a number of atomic experiments. IV. DIAGNOSTICS AND CHARACTERISTICS
In order to demonstrate the capabilities and characterize our 3D printed system, we testedits performance using spectroscopy and self-heterodyne analysis. For our first spectroscopyexperiment, we performed saturated absorption spectroscopy on a rubidium vapor.
Thissetup was used with the 3D printed system to examine the laser behavior and determine themode-hop-free tuning range and frequency resolution. Using the horizontal gross adjustment,the laser can easily be scanned over several nanometers and tuned to the region of atomictransitions.The observed transmission spectrum for the rubidium 5 S / to 5 P / transition for natural TABLE I. The parts required for construction of the 3D printed ECDL system. Possible retailerswith part numbers are supplied, but equivalent pieces can be found from a wide number of retailers.Part Specs Supplier and part number Price ($)Diffraction Grating 1800 lines/mm 12.7mm square Dynasil G1800R240CEAS 57Mirror square 12.7mm, < T r a n s m i ss i o n ( a r b ) Relative Frequency (MHz)
Rb F =2 g87
Rb F =3 g85
FIG. 2. A scan of the laser frequency over the rubidium isotope lines. This shows a scanning rangeof over 2 GHz. The system can also resolve the crossover peaks in Rb, with a separation of 32MHz. isotope abundance is shown in Fig. 2. While scanning the laser piezo, the 3D printed systemworks well for displaying the key features of saturated absorption. These data demonstratea mode-hop-free tuning range of over 2 GHz and the ability to resolve peaks separated by30 MHz.To gain further insight into the properties of the 3D printed system, it was also used at778 nm for two photon spectroscopy of the rubidium 5 S / to 5 D / transition. The laserwas found to scan without a mode-hop across hyperfine transitions for the different isotopes.A scan showing the peaks for the Rb F = 2 and Rb F = 3 transitions is shown in Fig.3. Due to the small hyperfine splitting in the 5 D state, this scan demonstrates the abilityto resolve peaks whose excitation frequency differs by only 8 MHz for the Rb, though thepeaks in Rb at 3 MHz to 5 MHz separations were not resolved.The system design was further tested using a self-heterodyne method in order toexamine the short-term linewidth. The laser was split, with half the beam going through an11 km delay line before recombination, in the traditional self-heterodyne setup. An AOMwas used in one of the paths to shift the center frequency by 68.5 MHz for easier observationin a spectrum analyzer. The spectrum was averaged over a 1 second time interval, with theresulting spectrum shown in Fig. 4. The linewidth of the laser was measured to be 1 . ± . .80.60.40.20.0 F l u o r e s c e n c e ( a r b ) Relative Frequency (MHz)
Rb F =2 g87
Rb F =3 g85
FIG. 3. Two-photon spectroscopy showing 420 nm fluorescence as a function of the laser frequencyas scanned over the 5 S / to 5 D / transition. Note the laser frequency separation is one half theground state hyperfine splitting for the two photon transition. number of research applications where extremely narrow linewidths are not necessary. It islikely that this linewidth could be further reduced through better protection of the systemfrom vibration and air currents. I n t e n s i t y ( a r b ) Relative Frequency (MHz)
FIG. 4. The spectrum for the auto-correlation observed using the self-heterodyne technique witha 1 second integration time. This gave a laser linewidth of 1 . ± . As most of the readily available 3D-printing materials are various types of plastics, thereis some concern about creep and degradation of the elastic properties of the system. In our8nvestigation, we have found that the integrity of the mount was maintained over the courseof weeks to months with only a minimal amount of adjustment to the system needed to keepit on an atomic resonance line, signifying that creep and degradation is not a hindrance ofperformance over that time scale. While this design has not been investigated with laserlocking, we expect that it should allow for standard locking mechanisms.
V. FUTURE DIRECTIONS
The laser housing design shown here has demonstrated tunability, stability, and adaptabil-ity for a large number of atomic applications. For those without access to metal machining,this 3D-printed design brings an ECDL to an achievable price point. While those with ma-chining ability may find cheaper and easier alternatives, 3D printing offers the ability forquick prototype turnaround and is widely available. As the state of the art in 3D printingbecomes more available, this model could be further adapted. The current form of this de-sign has demonstrated the ability to perform spectroscopy, and there are still advances thatcan be made. The print material and fill can be further optimized, as well as implementationon resin-based or metal printers. If further stabilization is necessary, adapting the design toallow for an airtight box surrounding the system, along with electrical connections that donot exert tension on the laser mount, would be helpful.Additional improvements include using a filament with high thermal conductivity to cre-ate a 3D-printed housing for the laser diode and thermoelectric cooler, eliminating the needfor the commercial Thorlabs diode mount. Though this modification would add additionalheat to the 3D printed material which may cause deformation, with careful planning of thegeometry of the print and multiple print materials these effects could be suppressed. Inaddition, different base and support print geometries can be scaled to suppress externalvibrations via phononic band gaps. CKNOWLEDGMENTS
We wish to acknowledge the contributions of Brandon Nelson to initial designs of theECDL, as well as Logan Hennes for testing adaptations to other wavelengths. ∗ [email protected] M. Fleming and A. Mooradian, “Spectral characteristics of external-cavity controlled semicon-ductor lasers,” IEEE Journal of Quantum Electronics , 44–59 (1981). K. B. MacAdam, A. Steinbach, and C. Wieman, “A narrowband tunable diode lasersystem with grating feedback, and a saturated absorption spectrometer for cs and rb,”Am. J. Phys. , 1098–1111 (1992). A. S. Arnold, J. S. Wilson, and M. G. Boshier, “A simple extended-cavity diode laser,”Rev. Sci. Inst. , 1236–1239 (1998). Lawrence N. Virgin, “Sympathetic resonance,” American Journal of Physics , 439–442 (2018),https://doi.org/10.1119/1.5025037. Jun Su, Weiguo Wang, Xu Wang, and Feng Song, “Simulation of the gravitational lensing effectof galactic dark matter halos using 3d printing technology,” Physics Teacher , 590 (2019). Lawrence Virgin, “Enhancing the teaching of linear structural analysis using additive manufac-turing,” Engineering Structures , 135 – 142 (2017). A. Heinrich, “Can one 3D print a laser?” in
Organic Photonic Materials and Devices XXII , Vol.11277, edited by Christopher E. Tabor, Franois Kajzar, and Toshikuni Kaino, InternationalSociety for Optics and Photonics (SPIE, 2020) pp. 1 – 7. The adjusted range of the final product allows for flexibility in the precision of the 3D printerused. Our choice was made due to availability. Any printer with a 4-inch by 4-inch bed couldbe used. C. J. Hawthorn, K. P. Weber, and R. E. Scholten, “Littrow configura-tion tunable external cavity diode laser with fixed direction output beam,”Review of Scientific Instruments , 4477–4479 (2001), https://doi.org/10.1063/1.1419217. Patrick McNicholl and Harold J. Metcalf, “Synchronous cavity mode and feedback wavelengthscanning in dye laser oscillators with gratings,” Appl. Opt. , 2757–2761 (1985). James S. Cybulski, James Clements, and Manu Prakash, “Foldscope: Origami-based papermicroscope,” PLOS ONE , 1–11 (2014). M. Saad Bhamla, Brandon Benson, Chew Chai, Georgios Katsikis, AanchalJohri, and Manu Prakash, “Hand-powered ultralow-cost paper centrifuge,”Nature Biomedical Engineering , 0009 (2017). Christopher J. Erickson, Marshall Van Zijll, Greg Doermann, and Dallin S. Dur-fee, “An ultrahigh stability, low-noise laser current driver with digital control,”Review of Scientific Instruments , 073107 (2008), https://doi.org/10.1063/1.2953597. K. G. Libbrecht and J. L. Hall, “A lownoise highspeed diode laser current controller,”Review of Scientific Instruments , 2133–2135 (1993), https://doi.org/10.1063/1.1143949. N. C. Pisenti, A. Restelli, B. J. Reschovsky, D. S. Barker, and G. K. Campbell, “Anultra-low noise, high-voltage piezo-driver,” Review of Scientific Instruments , 124702 (2016),https://doi.org/10.1063/1.4969059. Fabrizio Barone, Enrico Calloni, Aniello Grado, Rosario De Rosa, Luciano Di Fiore, LeopoldoMilano, and Guido Russo, “High accuracy digital temperature control for a laser diode,”Review of Scientific Instruments , 4051–4054 (1995), https://doi.org/10.1063/1.1145415. D. W. Preston, “Doppler-free saturated absorption: Laser spectroscopy,”Am. J. Phys. , 1432–1436 (1996). Abraham J. Olson, Evan J. Carlson, and Shannon K. Mayer, “Two-photon spectroscopy ofrubidium using a grating-feedback diode laser,” Am. J. Phys. , 218–223 (2006). T. Okoshi, K. Kikuchi, and A. Nakayama, “Novel method for high resolution measurement oflaser output spectrum,” Electronics Letters , 630–631 (1980). L. Richter, H. Mandelberg, M. Kruger, and P. McGrath, “Linewidth deter-mination from self-heterodyne measurements with subcoherence delay times,”IEEE Journal of Quantum Electronics , 2070–2074 (1986). Hang Xu, Amr Farag, and Damiano Pasini, “Routes to program thermal expan-sion in three-dimensional lattice metamaterials built from tetrahedral building blocks,”Journal of the Mechanics and Physics of Solids , 54 – 87 (2018). Kathryn H. Matlack, Anton Bauhofer, Sebastian Kr¨odel, Antonio Palermo, and hiara Daraio, “Composite 3d-printed metastructures for low-frequency and broadband vi-bration absorption,” Proceedings of the National Academy of Sciences