AVS 66th International Symposium & Exhibition
    Applied Surface Science Division Tuesday Sessions
       Session AS+BI+RA-TuM

Paper AS+BI+RA-TuM10
Extreme-Ultraviolet-Assisted Atom Probe Tomography

Tuesday, October 22, 2019, 11:00 am, Room A211

Session: Quantitative Surface Analysis III/Other Surface Analysis Methods
Presenter: Norman Sanford, National Institute of Standards and Technology (NIST)
Authors: N.A. Sanford, National Institute of Standards and Technology (NIST)
L. Miaja Avila, National Institute of Standards and Technology (NIST)
P. Blanchard, National Institute of Sandards and Technology (NIST)
D.R. Diercks, Colorado School of Mines
B. Gorman, Colorado School of Mines
A. Chiaramonti, National Institute of Sandards and Technology (NIST)
Correspondent: Click to Email

Laser-assisted atom probe tomography (L-APT) often returns compositional biases that deviate from a specimen’s true chemical makeup [1]. In our L-APT studies of GaN, which utilize a 355 nm laser, we find stoichiometric composition (50% Ga, 50% N) only when the pulse fluence is roughly 2E-7 J/cm2. Higher fluences return nonphysical, Ga-rich compositions; lower fluences give nonphysical N-rich compositions. L-APT of other materials, including SiO2, exhibit similar biases. L-APT is a thermally-activated mechanism. The specimen is subjected to a strong electric field that is held just below the threshold for field evaporation of ions; evaporation is triggered by thermal transients imparted by the incident laser pulses. Motivated by the need to improve the quantitative accuracy of atom probe tomography, we explored an alternative, photoionization-triggered pathway for field evaporation by replacing the conventional laser on an L-APT tool with a pulsed, extreme-ultraviolet (EUV) source—thus realizing the XAPT [2,3]. The tabletop EUV source is an ultrafast, amplified Ti:sapphire laser driving an Ar-filled capillary waveguide. Pulsed, coherent EUV is produced via high-harmonic generation [4]. The respective characteristics of EUV pulse width, repetition rate, wavelength (energy), and fluence, incident on a specimen tip are: 10 fs, 10 kHz, 30 nm (42 eV), and 2.5E-8 J/cm2. Comparative XAPT and L-APT analyses were performed on specimens of GaN and SiO2 (amorphous fused silica). For SiO2, L-APT returned an off-stoichiometry composition of 41% Si and 59% O; XAPT yielded 33% Si and 66% O (stoichiometry). For GaN, XAPT found 49% Ga and 51% N; L-APT yielded laser-pulse-fluence-dependent composition and returned 68% Ga and 32% N for a fluence of roughly 2E-6 J/cm2; L-APT conditions giving stoichiometric GaN appear above. The uncertainty of these measurements is 1—3 at. %. All specimens were examined at 50 K. XAPT-derived mass spectra show reduced occurrences of molecular ions and reduced “thermal tails” (persistent field emission following the laser pulse). Ongoing work for presentation at the meeting includes comparative studies of InGaN, Mg:GaN, and AlGaN. In all cases examined thus far, the XAPT approach offers measurable improvements over L-APT.

1. D. J. Larson, et al, Local Electrode Atom Probe Tomography (Springer, New York, 2013)

2. A. N. Chiarmonti, et al, MRS Advances (in press, 2019)

3. L. Miaja Avila, et al, Frontiers in Optic (FiO), Optical Society of America, Wash. DC, p. FTu5C. 3 (2018)

4. L. Miaja Avila, et al, Phys. Rev. Lett. 97, 113604 (2006)