Hydride Vapor-Phase Epitaxy

Hydride vapor-phase epitaxy (HVPE) and liquid-phase epitaxy (LPE) are older, lower-equipment-cost epitaxy methods that remain in active use for exactly the applications where their specific strengths — HVPE's very high growth rate, LPE's simplicity — outweigh MOCVD or MBE's superior interface control. HVPE grows thick GaN or III-V layers an order of magnitude faster than MOCVD; LPE grows a crystal layer directly out of a cooling metal solution with almost no specialized vacuum or gas-handling equipment at all.

Target: thick layer fast, or simple low-cost growth
Neither priority matches MOCVD/MBE's strengths
Method selection
HVPE
Hydride + halide vapor-phase reaction, very high growth rate
LPE
Crystal grows directly from a cooling saturated metal solution
Thick template layer, or simple low-cost crystal layer
Each method's specific strength delivered

HVPE's growth-rate advantage comes from running a fundamentally different, much faster chemical pathway than MOCVD uses. HVPE reacts a metal chloride vapor (made by passing HCl gas over molten gallium, for instance) with ammonia at the growth surface, a reaction that proceeds far faster than MOCVD's metal-organic decomposition chemistry — fast enough that HVPE can grow tens of microns of GaN in the time MOCVD grows a fraction of a micron. That speed is exactly why HVPE is the standard method for growing thick GaN template layers and freestanding GaN substrates, even though its interface and doping control can't match MOCVD's precision for the thin, carefully engineered device layers grown on top afterward.

HVPE and LPE — Older Methods, Still the Right Tool Sometimes speed and simplicity, traded against interface precision HVPE growth rate vs. MOCVD MOCVD ~µm/hr HVPE ~10s of µm/hr for thick template/substrate layers LPE — crystal from cooling solution molten metal solution cools, crystal precipitates on substrate no vacuum, no carrier gas needed both methods trade some interface control for speed or equipment simplicity

LPE's near-total lack of specialized equipment is its whole value proposition, and that's exactly why it persists in low-cost, lower-complexity applications. Growing a crystal layer by cooling a saturated metal solution next to a seed substrate needs no vacuum system, no precursor gas delivery, no plasma source — just a furnace, a crucible, and controlled cooling — which makes LPE dramatically cheaper to set up and operate than any vapor- or beam-based epitaxy method. The tradeoff is far less control over layer thickness, composition uniformity, and interface sharpness, which is why LPE has been displaced from most mainstream semiconductor production but still shows up where its cost advantage outweighs those limitations.

HVPE and LPE haven't survived as historical curiosities — they're still the right choice specifically when a process needs HVPE's raw growth speed for thick layers or LPE's minimal equipment footprint, and no amount of MOCVD or MBE's superior interface control changes that calculation for those particular jobs.

Take hydride vapor-phase epitaxy further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app