Unresolved intermediates in laser photodissociation of trimethylgallium: the carrier of the 215–220 nm metastable absorption bands, the $Ga_2^+$ surface channel, and bimodal photofragment desorption

Laser photodissociation of trimethylgallium (TMG, $Ga(CH_3)_3$), the archetypal gallium precursor for semiconductor thin-film growth, proceeds through short-lived neutral fragments whose identities control how absorbed UV photons are redistributed. A 1993 multiphoton-ionization time-of-flight study observed that a metastable neutral photofragment absorbs near 215 and 220 nm (one-photon), depleting the newly observed parent ion in the gas phase, and shows a corresponding two-photon resonance near 440 nm that enhances the $GaCH_3^+$ signal from 193-nm-photolyzed TMG multilayers on quartz at 110 K. The carrier was assigned inferentially to the methylgallium radical $GaCH_3$, with the dimethylgallium radical $Ga(CH_3)_2$ not excludable because its structure and spectrum were unknown. The problem is to replace this inference with direct assignments: measure the electronic spectra and structures of the candidate neutrals sufficient to settle the carrier of the 215–220 nm bands, assign the unexplained small $Ga_2^+$ surface signal, and establish the mechanism and coverage dependence of the bimodal (hot ~2000 K / cold ~260–320 K) translational-energy distributions of Ga and $GaCH_3$ desorbing from photolyzed TMG layers. Despite major adjacent progress — coverage-resolved GaAs studies in 1995, the first threshold-photoelectron spectrum of $GaCH_3$ (ionization energy 7.90 eV) in 2025 — no spectrum of neutral $GaCH_3$ in the relevant UV region and no structure of $Ga(CH_3)_2$ exist, so the problem remains open.

Background

Trimethylgallium (TMG, $Ga(CH_3)_3$) is the standard gallium-bearing organometallic precursor for metalorganic vapor-phase epitaxy (MOVPE, the growth of compound-semiconductor films from vapor-phase organometallic reagents) and a model system for laser-assisted, spatially selective deposition of gallium-containing films. Its ultraviolet (UV) photochemistry is therefore of both fundamental and practical interest; unwanted carbon incorporation from methyl groups is a known materials concern. UV photolysis of TMG proceeds by sequential loss of methyl groups, $Ga(CH_3)_3 \to Ga(CH_3)_2 + CH_3 \to GaCH_3 + CH_3 \to Ga + CH_3$, so the photodissociation is mediated by transient neutral radicals whose identity, structure, and optical properties determine how the absorbed photon energy is channeled.

The key experimental technique is multiphoton ionization time-of-flight (MPI-TOF) mass spectrometry, in which a pulsed laser supplies several photons that both dissociate and ionize the sample and the resulting ions are mass-analyzed by their flight times. When one absorbed photon is resonant with an intermediate electronic state of a neutral fragment (resonance-enhanced MPI, REMPI), the ion yield encodes an electronic spectrum of that neutral. Using one-color MPI-TOF with frequency-doubled dye-laser light at 209–227 nm (dye lasers emit tunable visible light whose second harmonic reaches the deep UV), Sato et al. (1993) observed the intact parent cation $Ga(CH_3)_3^+$ in the gas phase for the first time. Its yield showed pronounced, selective depletion near 215 and 220 nm, which the authors attributed to absorption by a short-lived neutral intermediate — a 'metastable' photofragment born after the first photon and able to absorb the ionizing photon within the same laser pulse. Mitchell, Hackett, Rayner and Humphries (1985) had shown that visible multiphoton dissociation of TMG produces a metastable molecular fragment, which they identified inferentially — from ion-yield behavior, without a spectrum — as $GaCH_3$ (at 222 nm it was produced to the exclusion of Ga atoms), and Zhang, Beuermann and Stuke (1989) had detected $GaCH_3$ by mass — but only as the cation $GaCH_3^+$ in far-UV laser mass spectrometry, never as a neutral with its own spectrum. The room-temperature-to-hot-gas UV absorption cross sections of TMG itself are known (Okabe et al. 1991), but not those of its photofragments.

The carrier question turns on a symmetry argument. If the absorbing molecule possessed an inversion center (a centrosymmetric, e.g. linear, geometry), one-photon and two-photon transitions would obey mutual exclusion (the Laporte parity rule), reaching states of opposite parity; the methylgallium radical $GaCH_3$ cannot be centrosymmetric, whereas a linear $Ga(CH_3)_2$ would be (a bent one would not). The 1993 study combined this parity reasoning with an energy-level diagram assembled from thermochemical and excitation data to argue that the carrier of the ~215–220 nm bands is most probably $GaCH_3$. The argument stays inconclusive because the molecular structure and electronic parity of the dimethylgallium radical $Ga(CH_3)_2$ were (and remain) unknown, and larger methylated fragments cannot be formally excluded either.

The same study photolyzed multilayers of TMG condensed on a quartz substrate held at 110 K with 193-nm light and probed the desorbing neutrals after a variable delay by MPI with the fundamental (undoubled) dye-laser output near 440 nm; two 440-nm photons carry the energy of one 220-nm photon, so a two-photon resonance there is the visible-wavelength signature of the same ~220 nm transition. The $GaCH_3^+$ signal indeed showed a broad band peaking near 440 nm, consistent with a resonant neutral precursor plausibly being $GaCH_3$; a small $Ga_2^+$ peak also appeared whose formation pathway was never assigned. Time-of-flight profiles of the neutral Ga and $GaCH_3$ fragments were bimodal and were modeled as sums of two Maxwell–Boltzmann (MB) distributions — the statistical velocity distribution of a thermalized ensemble, characterized by an effective temperature — with components near 2000 K (fast) and 260–320 K (slow, close to room temperature). Whether the fast and slow components arise from direct photodissociation of topmost adsorbed molecules, from subsurface recombination or ablation processes, and how all of this changes between monolayer and multilayer coverage, was left undetermined.

The broader context has since evolved substantially. Photodissociation of TMG on semiconductor surfaces was mapped in coverage-resolved detail on GaAs(100) (Zhu and co-workers, 1995), thermal decomposition networks of TMG in GaN growth have been analyzed with first-principles thermodynamics (Sekiguchi et al. 2017; Danielsson et al. 2016), and laser-induced fluorescence (LIF, detection of neutrals via their fluorescence after laser excitation) has very recently delivered the first spectra of TMG-derived radicals — GaCH$_2$ in 2024 — while the first direct spectroscopic characterization of $GaCH_3$ itself, a vibrationally resolved threshold photoelectron spectrum (a photoelectron method that detects only electrons ejected exactly at the ionization threshold, so the spectrum maps the ion's vibrational levels) giving its ionization energy (the minimum energy to remove an electron, IE), appeared only in 2025. None of these measurements, however, touches the neutral excited states near 215–220 nm that define the metastable in the photolysis experiments. The question that the 1993 study left open — what exactly are the transient species that carry the metastable absorption and mediate surface photofragment desorption — therefore still stands.

Problem Statement

The problem left open by the 1993 study of laser photodissociation of trimethylgallium (TMG) — and not closed by subsequent literature — is to replace the inferential identification of its transient intermediates with direct, spectroscopically grounded assignments:

(1) Carrier of the metastable absorption bands. Determine, by direct spectroscopy of the neutral species, which fragment absorbs near 215–220 nm (one-photon) — and correspondingly near 440 nm (two-photon) — in UV-laser photodissociation of TMG: the methylgallium radical $GaCH_3$, the dimethylgallium radical $Ga(CH_3)_2$, or another fragment. The assignment must be secure enough to exclude the competing candidates; since the inconclusiveness of the 1993 argument traces to the unknown structure and electronic parity of $Ga(CH_3)_2$, structural and parity information on this radical is directly relevant to its exclusion. Securing (or refuting) the one-photon/two-photon correspondence between the ~220 nm gas-phase depletion bands and the ~440 nm surface resonance is part of this item.

(2) The $Ga_2^+$ channel. Assign the formation pathway of the small $Ga_2^+$ signal observed in MPI probing of 193-nm-photolyzed TMG multilayers on quartz — for example, neutral $Ga_2$ formed by recombination at or below the surface, ion–molecule chemistry in the laser focus, or clustering of photofragments.

(3) Surface desorption mechanism and coverage dependence. Establish the mechanism producing the bimodal translational-energy distributions (Maxwell–Boltzmann components near 2000 K and 260–320 K) of neutral Ga and $GaCH_3$ ejected from photolyzed TMG multilayers on quartz at 110 K — in particular the relative roles of direct photodissociation of topmost adsorbed molecules versus subsurface recombination or ablation — and how the channels and their translational-energy partitioning change between monolayer and multilayer coverage.

These components share one objective: a mechanistic account of TMG UV photodissociation in the gas phase and in adsorbed layers in which the transient intermediates are identified through their own spectra and structures rather than through indirect energetic and symmetry arguments.

References

1. Akihiko Sato, Yugo Tanaka, Makoto Tsunekawa, Makoto Kobayashi, Hiroyasu Sato, "Laser photodissociation of trimethylgallium in the gas phase and on a quartz substrate as studied by multiphoton ionization-time of flight mass spectroscopy", The Journal of Physical Chemistry, 1993, vol. 97, pp. 8458–8463, DOI: 10.1021/j100134a015, https://doi.org/10.1021/j100134a015
2. Y. Zhang, Th. Beuermann, M. Stuke, "UV laser photofragment GaCH3 detected by far UV laser mass spectrometry", Applied Physics B, 1989, vol. 48, pp. 97–100, DOI: 10.1007/BF00694425, https://doi.org/10.1007/BF00694425
3. Th. Beuermann, M. Stuke, "Tunable UV laser photolysis of organometallics with product detection by laser mass spectroscopy: Trimethylaluminum", Applied Physics B, 1989, vol. 49, pp. 145–148, DOI: 10.1007/BF00332274, https://doi.org/10.1007/BF00332274
4. Th. Beuermann, M. Stuke, "Photolysis of group III (Al, Ga, In) trimethyl compounds: Detection of organic photofragments CH3 and C2H6 by picosecond laser mass spectroscopy", Chemical Physics Letters, 1991, vol. 178, pp. 197–203, DOI: 10.1016/0009-2614(91)87056-H, https://doi.org/10.1016/0009-2614(91)87056-H
5. W. Braun, R. Klein, A. Fahr, H. Okabe, A. Mele, "Laser photolysis of trimethylgallium at 193 nm: Quantum yields for methyl radical and ethane production", Chemical Physics Letters, 1990, vol. 166, pp. 397–403, DOI: 10.1016/0009-2614(90)85050-M, https://doi.org/10.1016/0009-2614(90)85050-M
6. S. A. Mitchell, P. A. Hackett, "Pulsed visible laser photolyis of B(C2H5)3, Al2(CH3)6, Ga(CH3)3, and In(CH3)3: Multiphoton ionization spectra of Al, Ga, and In atoms", The Journal of Chemical Physics, 1983, vol. 79, pp. 4815–4822, DOI: 10.1063/1.445627, https://doi.org/10.1063/1.445627
7. S. A. Mitchell, P. A. Hackett, "Visible multiphoton dissociation of gallium trimethyl: Time-resolved multiphoton ionization of gallium atoms", Chemical Physics Letters, 1984, vol. 107, pp. 508–514, DOI: 10.1016/S0009-2614(84)85148-9, https://doi.org/10.1016/S0009-2614(84)85148-9
8. S. A. Mitchell, P. A. Hackett, D. M. Rayner, M. R. Humphries, "Multiphoton dissociation of Ga(CH3)3", The Journal of Chemical Physics, 1985, vol. 83, pp. 5028–5042, DOI: 10.1063/1.449769, https://doi.org/10.1063/1.449769
9. Hideo Okabe, M. K. Emadi-Babaki, V. R. McCrary, "Temperature-dependent ultraviolet absorption spectra of group IIIb and Vb compounds used in photo-assisted chemical vapor deposition", Journal of Applied Physics, 1991, vol. 69, pp. 1730–1735, DOI: 10.1063/1.347219, https://doi.org/10.1063/1.347219
10. Akihiko Sato, Kazuo Kasatani, Hiroyasu Sato, "Photodissociation of trimethylgallium and trimethylaluminum on a quartz substrate as studied by laser induced fluorescence", Applied Surface Science, 1994, vols. 79–80, pp. 79–83, DOI: 10.1016/0169-4332(94)90391-3, https://doi.org/10.1016/0169-4332(94)90391-3
11. Akihiko Sato, Kazuo Kasatani, Hiroyasu Sato, "Laser photodissociation of trimethylaluminum and trimethylindium on a quartz substrate: Laser-induced fluorescence study", Applied Organometallic Chemistry, 1994, vol. 8, pp. 379–384, DOI: 10.1002/AOC.590080411, https://doi.org/10.1002/AOC.590080411
12. S. Cui, Donna K. Hacker, Q. Xin, R. A. Hinton, X.-Y. Zhu, "Photodissociation Dynamics of Trimethylgallium on GaAs", The Journal of Physical Chemistry, 1995, vol. 99, pp. 11515–11522, DOI: 10.1021/j100029a033, https://doi.org/10.1021/j100029a033
13. S. Cui, Donna K. Hacker, J. Pattanayak, Q.-S. Xin, R. A. Hinton, X.-Y. Zhu, "Photodissociation Dynamics of Trimethylgallium on GaAs", The Journal of Physical Chemistry B, 1997, vol. 101, p. 9844, DOI: 10.1021/jp971945e, https://doi.org/10.1021/jp971945e
14. Lon B. Knight, John J. Banisaukas, Robert Babb, Ernest R. Davidson, "Electron spin resonance matrix isolation and ab initio theoretical investigations of 69,71GaH2, 69,71GaD2, H69,71GaCH3, and D69,71GaCD3", The Journal of Chemical Physics, 1996, vol. 105, pp. 6607–6615, DOI: 10.1063/1.471974, https://doi.org/10.1063/1.471974
15. Gregory M. Greetham, Andrew M. Ellis, "Laser-induced fluorescence spectroscopy of the gallium dimer: evidence for a $^3\Pi_u$ electronic ground state", Journal of Molecular Spectroscopy, 2003, vol. 222, pp. 273–275, DOI: 10.1016/j.jms.2003.08.003, https://doi.org/10.1016/j.jms.2003.08.003
16. Riccardo Tarroni, Dennis J. Clouthier, "Ab initio spectroscopy of the aluminum methylene (AlCH2) free radical", The Journal of Chemical Physics, 2020, vol. 153, DOI: 10.1063/5.0010552, https://doi.org/10.1063/5.0010552
17. Fumie X. Sunahori, Tony C. Smith, Dennis J. Clouthier, "Spectroscopic identification and characterization of the aluminum methylene (AlCH2) free radical", The Journal of Chemical Physics, 2022, vol. 157, DOI: 10.1063/5.0101060, https://doi.org/10.1063/5.0101060
18. Tony C. Smith, Riccardo Tarroni, Dennis J. Clouthier, "Spectroscopic detection of the gallium methylene (GaCH2 and GaCD2) free radical in the gas phase by laser-induced fluorescence and emission spectroscopy", The Journal of Chemical Physics, 2024, vol. 160, DOI: 10.1063/5.0182504, https://doi.org/10.1063/5.0182504
19. Emil Karaev, Dorothee Schaffner, Marius Gerlach, Patrick Hemberger, Ingo Fischer, "Photoelectron Spectra of Small Gallium and Aluminum Radicals", The Journal of Physical Chemistry A, 2025, vol. 129, pp. 6004–6010, DOI: 10.1021/acs.jpca.5c02601, https://doi.org/10.1021/acs.jpca.5c02601
20. H.-T. Lam, J. M. Vohs, "Surface reactions of trimethylgallium on MOVPE-grown GaN(0001)", Surface Science, 1999, vol. 426, pp. 199–211, DOI: 10.1016/S0039-6028(99)00291-5, https://doi.org/10.1016/S0039-6028(99)00291-5
21. K. Sekiguchi, H. Shirakawa, Y. Yamamoto, M. Araidai, Y. Kangawa, K. Kakimoto, K. Shiraishi, "First-principles and thermodynamic analysis of trimethylgallium (TMG) decomposition during MOVPE growth of GaN", Journal of Crystal Growth, 2017, vol. 468, pp. 950–953, DOI: 10.1016/j.jcrysgro.2016.12.044, https://doi.org/10.1016/j.jcrysgro.2016.12.044
22. Örjan Danielsson, Xun Li, Lars Ojamäe, Erik Janzén, Henrik Pedersen, Urban Forsberg, "A model for carbon incorporation from trimethyl gallium in chemical vapor deposition of gallium nitride", Journal of Materials Chemistry C, 2016, vol. 4, pp. 863–871, DOI: 10.1039/C5TC03989D, https://doi.org/10.1039/C5TC03989D