Tolerance01
Orientation
Measured, not stamped. A Laue or XRD rocking curve per boule, referenced to the flat or notch on the actual part.
- Cut tolerance
- ±10′ (0.17°)
- Off-axis
- per your design
- Flat / notch
- to your datum
- Report
- XRD per lot
02Piezoelectric & scintillation
Crystals that behave
the way the datasheet
said they would.
Boules, blanks, plates and finished elements from growers we can trace to a single growth run. Orientation, dopant level and light yield go into the quotation as numbers — not into a phone call after delivery.
01Piezoelectric
Single-crystal piezoelectrics
Chosen for stability and low loss, not for raw displacement. If your design needs a large d constant you want PMN-PT; if it needs a frequency that does not drift, you want quartz — and the two are not interchangeable no matter what a trading company tells you.
Values below are typical room-temperature figures for the common cuts. Coupling and temperature behaviour change with cut angle, so the number that matters is the one measured on your lot.
Table 1 · Piezoelectric constants by material and cut
| Material | Common cuts | Piezo constant pC/N | Coupling k typ. | Curie / limit °C | Density g/cm³ | Typical use |
|---|---|---|---|---|---|---|
| α-Quartz SiO₂ | X, Y, Z, AT, ST, SC | d₁₁ 2.31 | 0.10 | 573 α→β | 2.65 | Resonators, delay lines, pressure & force sensors |
| Lithium niobate LiNbO₃ | Y/Z, 128° Y-X, X-cut | d₃₃ 6.0 d₁₅ 69 | 0.17–0.49 | 1,140–1,210 | 4.64 | SAW filters, EO modulators, ultrasonic transducers |
| Lithium tantalate LiTaO₃ | Y/Z, X-cut, 36° Y-X | d₃₃ 5.7 | 0.19–0.40 | 600–620 | 7.45 | Pyroelectric detectors, SAW, IR optics |
| Langasite La₃Ga₅SiO₁₄ | X, Y, Z | d₁₁ 6.7 | 0.16 | none to 1,470 melt | 5.75 | High-temperature sensors, downhole, aerospace |
| PMN-PT relaxor | 〈001〉 poled | d₃₃ 1,500–2,500 | 0.85–0.92 | 130–170 Tm | 8.1 | Medical ultrasound, actuators, sonar |
| AlN bulk & thin film | c-axis | d₃₃ 5.0–5.5 | 0.25 kt | >1,000 | 3.26 | HBAR, MEMS, high-temperature SAW |
Scroll the table sideways for all columns →
Note on k. The electromechanical coupling coefficient is a property of the cut, not of the material alone — the same LiNbO₃ boule gives a very different k on 128° Y-X than on Y-Z. We quote against your resonator design and ask the grower for measured impedance data, not a catalogue figure.
02Scintillation
Scintillator crystals
Light yield, decay time, density and hygroscopicity pull against each other. The right choice depends on whether you are counting photons slowly and precisely, or timing them in nanoseconds — so we start from your detector, not from a price list.
Table 2 · Scintillation properties, typical values
| Material | Light yield ph/MeV | Peak λ nm | Decay ns | Density g/cm³ | Hygroscopic | Where it wins |
|---|---|---|---|---|---|---|
| NaI(Tl) | 38,000–45,000 | 415 | 230 | 3.67 | Yes | General gamma spectroscopy; best resolution per dollar |
| CsI(Tl) | 52,000–56,000 | 550 | 680–1,000 | 4.51 | Slight | Rugged detectors, good match to photodiodes |
| CsI(Na) | 39,000–43,000 | 420 | 460–630 | 4.51 | Yes | Higher Z, faster than CsI(Tl), good blue match |
| BGO | 8,000–9,000 | 480 | 300 | 7.13 | No | High stopping power; PET, calorimetry, no packaging drama |
| LYSO(Ce) | 25,000–32,000 | 420 | 40 | 7.1 | No | TOF-PET and fast timing |
| GSO(Ce) | 8,000–10,000 | 440 | 30–60 | 6.71 | No | PET, well logging, high count rates |
| GAGG(Ce) | 46,000–58,000 | 520 | 90–100 | 6.63 | No | Bright and non-hygroscopic — the usual NaI(Tl) replacement |
| LaBr₃(Ce) | 61,000–70,000 | 380 | 16–35 | 5.08 | Yes | Best energy resolution and fast decay together |
| CeBr₃ | 57,000–66,000 | 380–390 | 18–20 | 5.10 | Yes | LaBr₃ performance without intrinsic activation |
| SrI₂(Eu) | 80,000–115,000 | 435 | 1,200 | 4.55 | Yes | Highest light yield available; isotope ID |
| CaF₂(Eu) | 19,000–30,000 | 435 | 950 | 3.19 | No | Large-area panels, low intrinsic background |
Scroll the table sideways for all columns →
Note on light yield. Published figures assume a specific size, dopant concentration and readout. A 1-inch cube and a 4-inch cylinder of the same CsI(Tl) do not produce the same number. Every quotation we send names the size the figure refers to, and we ask for the grower's own pulse-height spectrum on the lot being shipped.
03Written into the order
Three things we refuse to leave verbal
A crystal quotation that says "high quality, no bubble" is not a specification. These are the three headings we insist on filling with numbers before a PO is issued — and the same three we check against at pre-shipment.
Chemistry02
Purity & dopant
The difference between two "NaI(Tl)" quotes is usually the thallium concentration — and it moves the light yield by tens of percent.
- Trace metals
- ICP-MS
- Dopant level
- ±5% of target
- OH content
- quartz, on request
- Colour centres
- visual + spectra
Finish03
Surface & coating
Polish quality decides how much light ever reaches the photocathode. We hold the polish and the coating curve in the same document.
- Polished Ra
- ≤0.5 nm
- Parallelism
- ≤10″
- TTV
- ≤5 µm
- AR / HR
- curve supplied
04Qualification
How we qualify a grower
Six checks, done before the first paid order. On a new supplier we do them at the factory; on a repeat supplier we do them per lot and send you the file.
01
Boule traceability
The lot number must resolve to a growth run — furnace, date, seed — not to a warehouse bin. We ask to see the run log and match it to the parts we are buying.
02
Orientation by measurement
Laue back-reflection or XRD on the finished part, with the flat or notch in frame. A stamp on the box is not evidence; we have seen boxes stamped correctly and parts cut 2° off.
03
Phase and Curie point
For ferroelectrics, a dielectric peak or DSC trace near the stated Curie temperature. For quartz, a check for twinning under cross-polarised light — the defect that quietly halves Q.
04
Inclusion and bubble scan
Cross-polarised inspection against a ruled background, photographed. We count and locate, per part, against the acceptance number written in the PO — not "few, acceptable".
05
Transmission on the lot
A spectrophotometer curve from 200 to 2,500 nm on the material actually shipping. Catalogue curves are for catalogue material; absorption bands show up in the lot, not the brochure.
06
Packaging for hygroscopic stock
NaI(Tl), LaBr₃ and SrI₂(Eu) degrade in ordinary air. We verify vacuum sealing with desiccant, and we open and re-seal at pre-shipment inspection so you receive a seal that was closed after we looked.
05Where it goes wrong
Three failures we see most often
Trap 01
"Same grade" re-labelling
A dummy-grade boule is cut, polished and sold as detector grade because the invoice says so. The tell is usually the transmission curve that arrives as a photocopy of a photocopy, with no lot number on it.
Trap 02
Orientation as a stamp
Cutting tolerance is the first thing sacrificed when a factory is behind. Two degrees of error is invisible on a drawing and fatal on a resonator — and no one discovers it until the parts are already at your assembler.
Trap 03
The mount, not the polish
Most yield loss happens after the crystal is perfect: bonding, potting, thermal cycling. We ask for the assembly drawing early, because the finish and the flatness have to be chosen for the mount, not for the inspection report.
Other categories
Where the crystal ends up
03 · Wafers
Semiconductor wafers & substrates
Silicon, SiC and sapphire with TTV, bow and particle data attached.
View category 05 · HandlingWafer handling consumables
PFA cassettes, baskets and boxes that do not scratch what they carry.
View category All fiveOptics, ceramics & the rest
Precision optics and quartz glass, plus machined technical ceramics.
Back to overviewSend a drawing, or a spec sheet you were given.
We will come back with the grower, the grade we would accept, the numbers we would write into the PO — and the price with the factory invoice attached.