NaI Crystal Grower

Two sets of Kyropoulos (KY) crystal growers for NaI:Tl crystal growth (Figure I-18) were installed. Each set consists of a grower and an annealing furnace. A small test grower can charge about 2 kg of NaI powder and a large grower, which is for growing the real COSINE detector, can charge 150 kg of NaI powder (Table I-1). These growers were designed with Dr. Alain Iltis, a former employee of Saint-Gobain who made NaI crystals for the DAMA/NaI experiment.

Figure I-18: Kyropoulos crystal growers, KY01 (left) and KY02 (right)

Table I-1 : Characteristics of growers

NameMax temp’ZoneDimension of CrucibleMax loadRemarks
2 kg KY grower800 ℃3 zoneΦ120 mm × 135 mm2 kgTest grower
2 kg furnace800 ℃2 zoneAs-grown NaIAnnealing
150 kg KY grower800 ℃3 zoneΦ611 mm × 440 mm150 kgFor the COSINE
150 kg furnace800 ℃2 zoneAs-grown NaIAnnealing

NaI and TlI are very volatile materials when melted, and crystal growth has a tendency to push impurities back to melt, so it is difficult to make an optimum concentration of Tl(activator) and minimize Tl concentration difference between the top and bottom of a crystal. Conventional crystal growth methods produce a long height and a comparably narrow diameter (Figure I-19 (1)), so the Tl concentration difference between top and bottom is quite large. Our grower can grow short and wide-diameter (Figure I-19 (2)) crystals. This method can reduce concentration difference between top and bottom, so we can get quite uniform scintillation properties with our NaI:Tl crystals grown at CUP. For optimum concentration, we have to exercise the process several times to obtain good growing conditions and reduce growth time.

Figure I-19: Comparison of concentration between conventional growth method (1) and CUP’s growth method (2)

We are testing to grow small NaI crystals with the 2 kg Kyropoulos grower and have so far grown eight NaI crystals. Figure I-20 shows the grown NaI crystals, seven are pure NaI and one is 0.1 mol% Tl doped NaI crystal. There were minor issues like bubbles, a central hole, and contamination issues, but some of them were resolved. We need to improve growing conditions of NaI more. CUP-grown NaI crystals were measured with ICP-MS and results are summarized in Table I-2. KY01-1701 and 1702 were grown with the crystal grade NaI powder (Sigma Aldrich, >5N) and from KY01-1703 to 1804 were grown with the Merck NaI powder (4N). The ICP-MS result of KY01-1701 showed quite severe contaminations from quartz felt, so we no longer use quartz felt. In the rest of grown crystals, there are no observable contaminations. We can observe significantly good reduction of impurities in molybdate crystals for the AMoRE experiment but not for NaI crystals. Therefore, highly purified NaI powders are necessary.

Figure I-20: Grown NaI crystals

Table I-2: ICP-MS results of the grown NaI crystals

Raw powder
(Sigma-Aldrich, Crystal grade)
Quartz felt
(w/ quartz felt)
Seed crystal
Raw powder
(Merck, 99.99 %)

Figure I-21 shows XRD analysis results of four CUP-grown NaI crystals and all crystals are well crystallized.

Figure I-21: XRD analysis results of NaI crystals

The KY01-1703 and KY01-1804 crystals are cut and polished to measure scintillation properties. Polishing and measuring were done in a glove box in a nitrogen atmosphere. Figure I-22 shows pulse-height spectra of NaI crystals irradiated by 662 keV gammas from a 137Cs source. The light yield of a reference LYSO crystal is 33,000 ph/MeV, and the KY01-1703 (pure NaI) shows a light yield of ~3,000 ph/MeV. In case of the Tl doped NaI (KY01-1804) crystal, the light yield is ~40,000 ph/MeV. Typical Tl doped NaI shows about 40,000 ~ 50,000 ph/MeV.

We have practiced the technique many times to grow good quality NaI crystals. We expect that we soon can be ready for growing large NaI crystals with the 150 kg grower (KY02). Our goal with the large NaI crystal grower is to grow 115 kg of crystals with dimensions of Φ400 mm × 250 mm.