Introduction

The present stage of Indian nuclear program is based mainly on natural uranium dioxide-fuelled and heavy water-cooled pressurized heavy water reactors (PHWRs). A typical Indian PHWR consists of horizontally placed coolant channels consisting of pressure tubes (PTs). The PTs, at present in Indian PHWRs, are made of Zr-2.5 % Nb and are encircled by zircaloy-4 calandria tubes (CTs) separated with four garter spring spacers (Zr-2.5 Nb-0.5 Cu), which prevent these two tubes coming in contact with each other during their service lives. The PT carries the nuclear fuel elements engulfed in flowing high temperature and high-pressure heavy water coolant and undergoes nuclear irradiation [1]. It serves as the main pressure boundary of the reactor and its integrity is therefore central to the safety of a PHWR [2]. The inner surface of the PT continuously encounters corrosion reaction with the hot heavy water coolant flowing through it [3].

$${\text{Zr}}+{ 2}\;\left( {\text{H and D}} \right)_{ 2} {\text{O}} \to {\text{ZrO}}_{ 2} + { 4}\;\left( {\text{H and D}} \right)$$
(1)

Part of hydrogen and deuterium released from Reaction (1) are absorbed by the PT (termed as pick-up). Hydrogen isotopes (as H) below specified quantity do also remain as impurity during fabrication of the PT. Hydrogen isotopes (H and D) if present beyond the solubility limit (~60 ppmw at average reactor operating temperature of 573 K) can precipitate as brittle δ-hydride phase which can degrade the mechanical properties of the PTs due to hydride-induced embrittlement. This embrittlement can result in decreased gross fracture toughness of the material. The H and D may also lead to localized hydride blistering and/or delayed hydride cracking. The latter two phenomenon can take place well below the solubility limit of the hydrogen in the presence of localized stresses and cold spots [4, 5]. Hence, the measurement of hydrogen isotopes is important in quality control during manufacturing and life management of these components. In India, these tubes are manufactured at nuclear fuel complex (NFC), Hyderabad by hot extrusion of the billet followed by pilgering for cold work [6]. The initial hydrogen content is measured from the off-cuts of the manufactured tubes and has been specified a value of ≤5 ppmw of Heq (where Heq = H + ½D ppmw) [7]. After installation of PTs in various nuclear reactors, in situ slivering is carried out in some representative PTs at periodic intervals, as a part of safety assurance and life management, whereby about 100–200 μ thickness semi-lunar shaped sliver-samples are scraped from the inner surface of the PT by specially designed scraping tool and analyzed for their H and D concentration to assess H and D picked-up during operation [8]. Measurement at our laboratories is being normally carried out by two techniques namely differential scanning calorimetry (DSC) and hot vacuum extraction-mass spectrometry (HVE-MS). Measurement may also be verified by inert gas fusion (IGF) technique based H-analyser.

In DSC technique, the sample is heated up to 623–673 K and the temperature of maximum solid solubility of hydride is correlated to hydrogen concentration in the sample by applying the Arrhenius equation governing the solubility of H and D in zirconium [9]. The measurement equipment is pre-calibrated and validated by using certified standards of hydrogen in zirconium. DSC technique measures only the part of total H and D in the sample which is present as δ-hydrides/deuterides (at room temperature) in equilibrium with α-zirconium and quantifies only dissolving/precipitating H and D during heating/cooling cycles. If any part of H and/or D is trapped in the lattice defects or β-zirconium and does not dissolve/precipitate at TSSD/TSSP temperatures (623–673 K), those H and D will not be reflected in DSC measured Heq concentration [10].

In HVE-QMS technique, the sample is heated to 1353 K in vacuum (10−6 torr), H and D extracted from the metal matrix are collected in a collection flask having calibrated volume to estimate the total gas collected and passed through QMS to evaluate H and D in the sample [11, 12].

In IGF technique, the sample is heated to near its melting point (>1873 K) under inert carrier gas (argon/nitrogen) to extract H and D present in the sample and passed through thermal conductivity detector for estimation of Heq concentration [13]. As the thermal conductivity of H2, D2 and HD are different; the system has to be calibrated by standard gas having same isotopic composition as the sample.

The DSC and IGF techniques measure the total content of H+D, represented as Heq without differentiating between hydrogen and deuterium while the HVE-MS technique is capable of measuring the H and D separately and representing the Heq by adding them up as (H + ½D ppmw). DSC technique relies on the values of TSS and is therefore a secondary measurement [14].

It has been observed that the reported results of measurements of Heq invariably differ with the technique being used. This can be attributed to the differences in the principles and experimental parameters of the techniques (like maximum temperature, applied vacuum etc.) and the binding nature of H and D in the matrices of the sample [15]. Hence, the extraction parameters of H and D isotopes from different zirconium alloys which could be responsible for the ambiguity in the previously observed results needs to be studied to establish a clear understanding of the results being obtained from different techniques and apply for correction/standardization of the analytical parameters (if required). In this context the present work reports evolution profile of hydrogen isotopes in zircaloy-2, zircaloy-4 and Zr-2.5 % Nb as a function of temperature which could be a major parameter influencing the extraction and consequently estimation of hydrogen (and its isotopes) by presently used techniques. The experiments have been carried out by thermo-gravimetric analysis (TGA) coupled with evolved gas analysis (EGA).

Experimental

Sample preparation

Rectangular coupons generated from zircaloy-2, zircaloy-4 and Zr-2.5 % Nb sheets were charged with hydrogen. Deuterium charged samples will be prepared in the next stage of experiments. The chemical composition of the coupons is as per specification for clad and pressure tube materials used in Indian BWRs and PHWRs. The coupon details are given in Table 1. Initial hydrogen content in the coupons was analysed by IGF technique based hydrogen analyser.

Table 1 Details of the coupons used for hydrogen charging
Full size table

Hydrogen charging

The coupons were electrolytically charged with hydrogen in alkaline medium (0.1 % solution of NaOH and As2O5) at constant current density of 0.1 A/cm2 at 298 K. Samples from charged coupons were then cut by slow speed diamond cut off machine, cleaned and analyzed by IGF technique for their bulk hydrogen content.

Sample preparation for TGA-EGA

Samples in the mass range of 190–220 mg were generated from the hydrogen charged coupons by slow speed cut off machine. The coupon was cut at a speed of 90–100 RPM and CCl4 was used as the coolant during cutting to avoid any hydrogen pick up during the process. The samples were cleaned in acetone and dried at 333–343 K and used for TGA-EGA analysis.

TGA-EGA experiment details

For the present study, commercial thermoanalysis system Labsys Evo procured from M/s Setaram Instrumentation coupled with quadrupole mass spectrometer (QMS) Omnistar from M/s Pfeiffer Vacuum was used (Fig. 1). This set-up allows for the continuous recording of mass change and simultaneous control of the gaseous environment during the experiment. The high temperature furnace of the thermobalance is designed for maximum temperatures of 1873 K. The QMS is capable of measuring hydrogen concentration in the range of sub ppm by weight of the sample. The system was calibrated for mass and temperature by weight loss of copper pentahydrate and Au, Ni certified standards respectively at heating rate of 10 K/min and HP Ar flow rate of 30 mL/min. The prepared samples were weighed and placed in the alumina crucible placed in the hot zone of the furnace. The sample was then heated up to 1873 K in alumina crucible at a programmed heating rate of 10 K/min and HP Ar flow of 30 mL/min. The sample was held at 1873 K for 5 min before cooling to room temperature. HP Ar initially 99.998 % pure was further purified by putting additional moisture and oxygen traps in the carrier gas line. Thermo-grams were recorded and the gases evolved from the sample were simultaneously analyzed by QMS. A blank run with empty crucible was also taken. Samples charged with hydrogen were analyzed in the present set of experiments. However, hydrogen and deuterium having similar chemical properties are expected to have similar evolution profiles. The deuterium charged samples will be taken in the next phase of the studies. Hence, the present discussion will be based on results on hydrogen evolution profiles from different alloys. The discussion is however expected to hold true for deuterium also.

Fig. 1
figure 1

Schematic of experimental set up

Full size image

Results and discussion

The evolution profile of hydrogen from the sample is shown in Fig. 2. The broken line represents the weight change (in mg) and the solid line indicates the hydrogen ion current (A) due to hydrogen gas getting evolved from the sample as a function of temperature. Slight weight loss in thermogram up 800 K could be attributed to loss of adsorbed volatiles on the sample surface. Increase in weight due to oxidation of the sample by alumina crucible was observed after 800 K. Increase in weight could be observed in all the samples. The oxidation is parabolic initially followed by a linear rate after 1073 K as expected for zirconium based matrices. Change in weight of the sample thus could not be evaluated to quantitatively estimate the evolved hydrogen. The blank run with empty crucibles was taken and the recorded thermogram was subtracted from subsequent curves to nullify the buoyancy effect. The hydrogen evolution profile in the blank run showed decrease in the hydrogen ion current followed by no change at higher temperatures. The hydrogen evolution curves were similar in nature for zircaloy-2, zircaloy-4 and Zr-2.5 % Nb. Similar sample weight and surface/volume ratio was maintained for all the alloys to nullify the effect of sample dimensions on evolution. All the alloy samples indicate one minor evolution of hydrogen near 850–900 K followed by a major evolution 1873 K. The major evolution starts before the minor evolution ceases. Hence, there is no marked separation between the two peaks. The aim of the present work is to estimate the temperature range over which the evolution of hydrogen from different zircaloy matrices takes place. The two step evolution does not affect the results of the present paper. For simplicity, we will be considering the onset temperature of evolution as the start of the minor evolution and the peak temperature of evolution as the peak of major evolution from the matrix. The results have been tabulated in Table 2. The peak temperatures of evolutions were observed to be 1540, 1564 and 1442 K for zircaloy-2, zircaloy-4 and Zr-2.5 %Nb respectively. Zr-2.5 Nb having the highest concentration of hydrogen shows lowest peak temperature of evolution. To assess whether complete evolution of hydrogen has taken place from the samples, residual hydrogen concentration was measured by IGF technique 4. Zr-2.5Nb samples showed lesser residual hydrogen as compared to Zircaloy-2 and Zircaloy-4. Hence, the percentage of hydrogen evolved was higher in Zr-2.5 % Nb.

Fig. 2
figure 2

Evolution profile of hydrogen in zircaloy-4

Full size image
Table 2 Onset and peak maximum temperature observed in TGA-EGA
Full size table

The difference in the onset and peak maximum temperature of H evolution can be attributed to the presence of different alloying elements in the zirconium matrix. The mechanism of absorption and desorption of hydrogen isotopes has been widely studied. Different models have been proposed to explain the evolution process of hydrogen in zirconium alloys but the exact mechanism of the process is still not established [16, 17].

Evolution of hydrogen from metal matrices takes place in two steps. In the first step δ-hydride/deuteride dissociates to form Zr and H and D dissolved in Zr matrix. In the second step, this dissolved H and D then diffuses to the surface from the solution phase and gets desorbed into the gas phase. The atomic H and D that is desorbed from the matrix recombines to form molecular hydrogen, HD and deuterium respectively. According to the kinetic model suggested by Terrani et al., the rate-controlling step in the desorption process is the recombination of H and D atoms on the surface to form H2, HD and D2 [17]. In another model suggested by Lauger et al., the activation energy required for evolution of hydrogen from the zirconium matrix is dependent upon the heat of solution of atomic hydrogen in the zirconium matrix. It is a well known fact that alloying elements can influence the solubility of hydrogen in zirconium matrix. However, before affirming the role of alloying elements the mechanism of evolution of hydrogen needs to be established.

Conclusions

Evolution profile of hydrogen isotopes from zirconium based matrices is reported up to 1873 K using TGA-EGA. The peak maximum temperatures of evolution of hydrogen seem was observed to differ in zircaloy-2, 4 and Zr-2.5 % Nb and seem to depend upon the alloying elements in zirconium matrix. The mechanism of desorption of hydrogen isotopes needs to be studied in further detail to establish the role of alloying elements on desorption process.