Can Cr(III) substitute for Al(III) in the structure of boehmite?

Abstract

The dissolution of boehmite is a technical issue for Al industry because of its recalcitrant nature. In fact, a similar problem exists with boehmite in nuclear waste sludge at the Hanford site in Eastern Washington State, USA. Dissolution of Al phases is required to reduce the waste loadings in the final borosilicate glass waste form. Although not the most common Al-bearing species in the sludge, boehmite may become a rate limiting step in the processing of the wastes. Hanford boehmite is an order of magnitude more resistant to dissolution in hot caustic solutions than expected from surface-normalized rates. We are exploring potential intrinsic and extrinsic effects that may limit boehmite reactivity; one clue comes from microstructural analyses that indicate an association of Cr with Al in the Hanford nuclear waste. Hence, in this first paper, we investigated the potential role of chromium on the reactivity of boehmite in caustic solution. An important finding was that irrespective of the synthesis pathway, amount of Cr(III), or the resultant morphology, there was no evidence for Cr incorporation in the bulk structure, in agreement with QM calculations. In fact, electron microscopic (EM) and spectroscopic analyses showed that Cr was enriched at the (101) edges of the boehmite. However, Cr had no measurable effect on the morphology during the synthesis step. In contrast, comparison of the morphologies of the synthetic Cr-doped and pure boehmite samples after exposure to caustic solutions provided evidence that Cr inhibited the corrosion. TEM showed that Cr was not homogeneously distributed at the surface. Consequently, Cr may have partially passivated the surface by blocking discrete energetic sites on the lateral surfaces of boehmite.

---

Introduction

Boehmite (orthorhombic γ-AlOOH) is among the most abundant natural ores of aluminum, and occurs as a secondary mineral in weathered surface zones in limestones and clays as well as “silica-poor” igneous rocks. It can be bulk-produced with a wide range of desired functionalities and morphologies,^1–3 and has found a broad range of applications in industrial processes.^4–6 Boehmite aggregates can have a large surface area with both amphoteric functional groups on lateral faces and a stable protonated basal (010) surface that can be differentially expressed. Consequently, boehmite is a versatile adsorbent over a broad range of pH for both anions and cations.^7,8 The Bayer extraction process of alumina from ores utilises caustic leaching at elevated temperatures and pressures. The most common Al-bearing minerals found in bauxite are gibbsite (Al(OH)₃) and boehmite. Gibbsite dissolves relatively easily but boehmite is extremely recalcitrant in caustic solutions. This resistance to dissolution can be exacerbated by the presence of other phases, such as anatase (TiO₂). Ireland et al., using electron microscopy, demonstrated that the reaction with NaOH resulted in the formation of a mixed Al–Ti oxide layer that inhibited further reaction.³ It is likely that other contaminants might also inhibit boehmite dissolution in a similar manner.

Boehmite is also a common material found in the radioactive waste storage tanks at the Department of Energy (DOE) Hanford site in Eastern Washington State, U.S.A.⁹ Hanford was one of the major U.S. sites of Pu weapons production from 1943 until 1989. As part of the process of extracting Pu from irradiated U metal fuel, large quantities of liquid and solid wastes were produced. These wastes were stored in 177 steel tanks ranging in size from 35 000 gallons to one million gallons – all awaiting final processing into vitrified borosilicate glass logs which will finally be stored in a geologic waste repository.⁹

The wastes are a highly complex mixture of phases; including Al-oxides, iron oxides, chromium phases, sodium uranium oxides, and other phases and species in a high pH solution.¹⁰ The sludge also has a highly variable particle size range. Much of the sludge solids have formed heterogeneous agglomerates.

The Al in these wastes originated from one of three sources: dissolved Al-nuclear fuel cladding, Al(NO₃)₃ that was added to waste to increase the ionic strength during processing, and Al that was added as a corrosion control agent for the steel tanks.^11,12 The solubility limiting phase in the Hanford tanks is believed to be gibbsite. However, temperatures in the tanks briefly became elevated (boiling) due to radioactive decay from ⁹⁰Sr and ¹³⁷Cs; this resulted in conversion of some of the gibbsite to the more recalcitrant phase, boehmite. The radioactively hot elements were subsequently removed and since this time, the sludge temperature has not been greater than 45 °C.¹³ Although necessary for glass formation, too much Al results in a lower quality waste form, so efforts have been concentrated on removing a large fraction of the Al waste in the sludge. Studies have shown that boehmite in Hanford tank waste dissolves by a factor of ∼40× slower than predicted for surface area normalized rates. An outstanding question is whether this is due to extrinsic factors such as agglomeration reducing reactive surface area or intrinsic factors such as incorporation or adsorption of stabilizing elements.

In this study, we examined the potential role of chromium on limiting boehmite reactivity under caustic conditions. Cr(III) is a major species in the tank wastes.¹⁴ Indeed, Cr has been observed to be associated with Al in REDOX type wastes; however, there is no direct evidence for incorporation of Al and Cr in the same phase.^15,16 Other studies have implied that up to 20% Cr(III) could be incorporated into boehmite nanofibers under hydrothermal conditions.¹⁷ Further, increasing Cr concentrations systematically raised the dehydroxylation temperature of boehmite suggesting that Cr might increase boehmite stability, which could account for boehmite's recalcitrant behavior during caustic leaching of Hanford tank waste. In this contribution, we take the first steps towards understanding the interaction of Cr(III) and boehmite under conditions that are relevant to the Hanford tanks. First order questions are: (1) Is Cr(III) incorporated into the bulk structure of boehmite or adsorbed to the surface? (2) Does the morphology of boehmite or synthesis conditions affect the mode of association of Cr(III) with boehmite? (3) Is Cr(III) a determinant of boehmite morphology during synthesis? and (4) Does Cr(III) affect the morphology of boehmite under caustic leaching conditions relevant to tank waste processing conditions? We investigated two different reaction pathways and starting reactants. One pathway is the single step alkaline dehydration of gibbsite to rhombic boehmite plates that are similar to, albeit larger than, boehmite in the Hanford tanks,^15,18 (referred to here as synthesis 1). The other pathway is a two-step process that begins with the alkaline hydrolysis of Al(NO₃)₃ to Al(OH)₃, followed by the dehydration step, and yields elongated two-dimensional platelets (i.e., laths; referred to here as synthesis 2). Reaction products were characterized by electron microscopy and various spectroscopic methods. Experimental results were given context by 0 K ab initio calculations of the incorporation energies of Cr(III) substitution for Al(III) in the boehmite structure as a function of Cr–Cr separation and contrasting the bulk to near-surface environments.

Results and discussions

We synthesised two differently shaped boehmite particles; rhombohedral plates and nano-laths, to study the structural effect on Cr incorporation rate. Throughout this article the plates and laths will be referred to as synthesis 1 and 2 respectively, with Cr doping annotated as Cr_a and Cr_b for 1 and 10 atomic wt%. As discussed in the experimental section we characterised these particles using electron microscopy (EM), X-ray structural analysis (XRD), neutron diffraction (ND) and various spectroscopic techniques. The experimental data were then compared to theoretical calculations.

Diffraction

X-ray diffractograms of the undoped and Cr(III) doped boehmite samples are shown in Fig. 1a. The diffractogram of 1 and 2 indexed as pure boehmite with an orthorhombic unit cell (a = 3.700 Å, b = 12.227 Å, c = 2.868 Å, space group symmetry C_mcm, JCPDS PDF no. 21-1307).¹⁹ The sharp, intense peaks indicate high crystallinity. No distinct Cr containing phases were detected for the Cr containing samples. Further, XRD did not detect any effect from changing the synthetic conditions and consequent morphologies (compare XRD of 2 and 2-Cr_a to 1 and 1-Cr_a).

Fig. 1 (a) X-ray and (b) neutron diffractograms of boehmite and Cr(III)-doped boehmite composites: (a) (black trace) undoped boehmite 1; (red trace) 1% Cr(III)-doped boehmite 1-Cr_a; (blue trace) 10% Cr(III)-doped boehmite 1-Cr_b; (orange trace) undoped boehmite 2. (Brown trace) 1% Cr(III)-doped deuterated boehmite 2-Cr_a. (b) Colored circles represent the respective observed intensities while the black traces represent the calculated values. The green traces represent the difference. The prominent Cr₂O₃-like phase in 1-Cr_b is highlighted with an asterisks. The purple vertical bars represent the diffraction patterns of boehmite and Cr₂O₃ reported in literature.

The neutron diffraction (ND) results (Fig. 1b) are consistent with the X-ray diffraction measurements.²⁰ The diffractogram of 1-Cr_a looks nearly identical to the undoped sample with no separate or distinct Cr phase detected. Further, a Rietveld analysis assuming complete incorporation (i.e., 1%) of Cr in the octahedral Al(III) sites does not match the experimental pattern, strongly suggesting that a portion of Cr must be external to the boehmite structure. In fact, at higher Cr concentration (1-Cr_b), while boehmite is still the dominant phase, an additional peak appears that can be indexed to Cr₂O₃. Similar to XRD, no changes are detected as a function of the synthesis pathway (compare patterns for 1 and 2).

In summary, neither XRD nor ND provide compelling evidence for incorporation of Cr(III) into the boehmite structure. In fact, Rietveld refinement of the ND data suggests that Cr concentrations as low as about 0.1% in the boehmite structure should be detectable. This suggests that multiple Cr(III) environments are possible when the Cr concentration exceeds 0.1%, as is the case here. The following spectroscopic analyses largely confirm this finding.

Spectroscopy

Infra-red spectroscopy. The vibrational spectrum of boehmite sample 1 showed bands at 3275, 3080, 1189, 1068, 732, 608, and 465 cm⁻¹ (Fig. 2a), which agree with those reported in the literature.²¹ The IR spectra of 1-Cr_a and 1-Cr_b are nearly identical to that of the undoped species, showing no discernible changes and suggests that addition of Cr(III) did not affect the bulk vibrational structure of AlOOH. For samples 1, 1-Cr_a and 1-Cr_b, two broad bands at ca. 3275 and 3080 cm⁻¹ are observed symmetrically disposed relative to one another and are assigned to the ν_as(Al)O–H and ν_s(Al)O–H stretching vibrations.²² The intense band at 1068 cm⁻¹ and the shoulder at 1189 cm⁻¹ in 1, 1-Cr_a and 1-Cr_b are ascribed to the δ_s Al–O–H and δ_as Al–O–H bending vibrations in the boehmite lattice. For all the samples, the three strong bands observed at 732, 608 and 465 cm⁻¹ can be assigned to the vibration modes of AlO₆. The different synthetic routes and morphologies had no effect on the bulk vibrational properties of boehmite, as illustrated by the near identical spectra of 1 and 2. Additionally, similar to 1-Cr_a, Cr(III) had no detectable effect on the bulk vibrational properties of the nano-lath morphology (compare 2 to 2-Cr_a).

Fig. 2 (a) IR spectra of boehmite and Cr(III)-doped boehmite composites: (black trace) undoped boehmite 1; (red trace) 1% Cr(III)-doped boehmite 1-Cr_a; (blue trace) 10% Cr(III)-doped boehmite 1-Cr_b; (orange trace) undoped boehmite 2; (brown trace) 1% Cr(III)-doped deuterated boehmite 2-Cr_a, (b) and (c) Raman spectra of (black trace) undoped boehmite 1 and (orange trace) undoped boehmite 2.

Raman spectroscopy. The Raman spectrum of the boehmite sample 1 shows four weak and broad bands which are the distinguishing features for boehmite (Fig. 2b and c). Two broad bands at 3220 and 3080 cm⁻¹ are assigned to the symmetric and asymmetric stretches of the hydroxyl groups in the bulk material,²² and presumably correspond to the 3275 and 3080 cm⁻¹ infrared absorption bands. In addition, they have shoulders at 3305 and 2989 cm⁻¹ that have previously been assigned to surface hydroxyl groups.²² The low wavenumber region (200–1400 cm⁻¹) consists of hydroxyl translation modes characterized by strong bands at 453, 497 and 677 cm⁻¹ and a weaker band at 732 cm⁻¹. Addition of Cr(III) to the system had no discernible effect on the Raman spectra.

Photoelectron spectroscopy. The XPS data is listed in Table 1 and spectra are presented in Fig. 3. The O/Al and hydroxyl : oxide ratios for the non-doped boehmite (sample 1) are 2.01 and 1.00 which are consistent with expected stoichiometric ratios. The Al2p and O1s (oxide and hydroxyl) binding energies are higher and lower, respectively, than reported previously by Kloprogge et al.²³ However, there is no information given on the boehmite morphology or synthetic route in Kloprogge et al.; given that boehmite crystallographic faces can be differentially expressed by changing the synthesis conditions, it is hard to make comparisons between the two studies. Regardless, our interest is in differences between the Cr and non-Cr syntheses, where the morphologies and pH were identical. In this regard, the binding energy for Al2p drops slightly (−0.37 eV) in 1-Cr_a relative to the “undoped” sample 1. Conversely, the O1s binding energy increases by +0.87 eV for the oxide peak and +0.89 eV for hydroxyl. A peak at 533.46 eV indicates a minor physi-sorbed H₂O component. Therefore, relative to the marginal change in Al binding energy, the change for O is large and in the opposite direction suggesting the increase in O binding energy is real. Whether this variation in O1s binding energy and the increase in the OH/O²⁻ ratio is related to Cr or the sharp increase in the carbon overlayer thickness cannot be determined presently. Whereas Cr(III) is marginally more electronegative than Al(III), which is consistent with a higher O1s binding energy, the increased carbon overlayer thickness decreases the information depth such that more of the signal originates from the very near-surface.

Table 1 Tabulation of XPS data

Sample	Elemental form	B.E. (eV)	FWHM (eV)	Ratios
				O/Al	Al/C	Cr/Al
a C–O calculated from C1s line.b Chemisorbed water.c Physisorbed water.d There was a significant contribution to the O from C–O species, which was subtracted out.
1	Al2p	74.58	1.65	2.01	3.2	NA
	O1s (100%)	529.85	2.74
	O^2− (47.9%)	529.43
	OH^− (52.1%)	530.71
	C–O^a (1.5%)	n.d.
1-Cra	Al2p	74.21	1.43	2.07	1.9	0.020
	O1s (100%)		2.71
	O^2− (44.8%)	530.30
	OH (53.5%)	531.60
	H2O^b (1.7%)	533.46
	C–O^a (3.4%)	n.d.
	Cr2p3/2	576.34	2.65
1-Crb	Al2p	74.29	1.46	1.91(1.73)^d	1.0	0.022
	O1s	n.a.	2.73
	O^2− (36.5%)	530.43
	OH (48.2%)	531.65
	H2O^b (3.8%)	533.82
	H2O^c (3.8%)	535.20
	C–O^a (9.6%)	532.49
	Cr2p3/2	576.46	2.87

---

Fig. 3 Photoelectron spectrum of 1% Cr doped boehmite, 1-Cr_a.

The binding energy for Cr2p_3/2 is about −0.7 eV lower than expected for Cr(OH)₃ but similar to that for Cr₂O₃.²⁴ The peak shape, however, lacks the prominent multiplet structure characteristic of Cr₂O₃. Nonetheless, the Cr/Al ratio is twice the expected stoichiometric ratio, which indicates that Cr is not homogeneously incorporated in the bulk boehmite, but rather concentrated towards the near-surface.

The XPS of 1-Cr_b is similar to 1-Cr_a with the following exceptions: there is a greater contribution from both chemi- and physi-sorbed H₂O, and the O/Al ratio is lower. These differences could be due to the increased Cr over-layer in sample 1-Cr_b that increases the proportion of signal from the very top surface. Further, it is curious that the Cr/Al ratio is only 10% higher than for 1-Cr_a, given that the stoichiometric ratio is 10 fold higher. A consideration is that the Cr/Al and O/Al ratios will be affected by the over lay Cr due to the fact that Cr2p and O1s emit lower energy photoelectrons with smaller mean free paths than Al2p. Consequently, for the same absolute Cr and Al concentrations, the thicker the C over-layer the lower the Cr/Al and O/Al ratios. Alternatively, a greater proportion of Cr might be tied up in precipitates where the bulk of the Cr is not detected by XPS due to its surface sensitivity. In this regard, that neutron diffraction (ND) detected a Cr₂O₃-like phase in sample 1-Cr_b yet XPS does not provide compelling evidence for this phase, is an indication that XPS is averaging over more than one Cr(III) bonding environment, whereas ND is highlighting the crystalline phase. Evidence for more than one Cr bonding environment is provided in the following section on emission spectroscopy.

Cr luminescence emission spectroscopy. The Cr emission spectra of 1-Cr_a, 1-Cr_b and 2-Cr_a show nearly identical spectral profile (Fig. 4a). The spectra are characterized by two sharp bands at 688 and 701 nm, with shoulders at 715 and 725 nm. The lifetime and intensities are consistent with d–d transitions arising from a Cr(III) center. The overall emissive relaxation process can be fit with two distinct decay rates which is indicative of multiple relaxation pathways arising due to at least two different Cr(III) coordination environments.

Fig. 4 (a) Time resolved Cr(III) luminescence spectra of 1-Cr_a as a function of time: 20 μs (violet top-most trace), 40 μs (blue trace second from top), 60 μs (dark green trace third from top), 80 μs (dark green trace third from top), 100 μs (light green trace fourth from top) spectrum at, 120 μs (yellow trace fourth from bottom), 140 μs (orange trace third from bottom), 160 μs (red trace second from bottom), 180 μs (brown bottom-most trace). (b) Steady state luminescence spectra and (c) emission decay of boehmite and Cr-doped boehmite composites upon 400 nm excitation: (red trace) 1% Cr doped boehmite, 1-Cr_a; (blue trace) 10% Cr doped boehmite, 1-Cr_b; (green trace) 1% Cr doped boehmite, 2-Cr_a.

The luminescence relaxation profiles of 1-Cr_a, 1-Cr_b and 2-Cr_a were plotted as a function of time (Fig. 4c), where symbols represent experimental emission intensity, and lines are the bi-exponential fit of the emission intensity decay. The bi-exponential decays of the intensities can all be represented by Ae^−Bt + Ce^−Dt; the decay constants being shown in Table 2. The decay equations for each case are as follows, where I_E is emission intensity:

Table 2 Tabulation of emission decay constants

Boehmite species	Emission decay constants
	A	B	C	D
1-Cra	1110 ± 9	0.00021 ± 0.00002	32 549 ± 175	0.1289 ± 0.0007
1-Crb	1418 ± 67	0.003321 ± 0.00059	52 494 ± 319	0.1528 ± 0.0012
2-Cra	1210 ± 67	0.00152 ± 0.00091	42 658 ± 259	0.1378 ± 0.0001

---

1-Cr_a:

I_E = (1110 ± 9)e^{−(0.00021±0.00002)t} + (32 549 ± 175)e^{−(0.1289±0.0007)t} (1)

1-Cr_b:

I_E = (1418 ± 67)e^{−(0.00332±0.00059)t} + (52 494 ± 319)e^{−(0.1528±0.0012)t} (2)

2-Cr_a:

I_E = (1210 ± 67)e^{−(0.00152±0.00091)t} + (42 658 ± 259)e^{−(0.1378±0.0009)t} (3)

Closer inspection of the luminescence relaxation profile of 1-Cr_a (Fig. 4c) indicates that the transitions at 688 and 701 nm (14 535 and 14 265 cm⁻¹ respectively) are very short lived (k = 0.1289 ± 0.0007 μs⁻¹). These are tentatively assigned to the 0, 0 components of the ²E → ⁴A₂ transition (in O symmetry) split by a trigonal field (C₃ symmetry).²⁵ The spectra are indicative of isolated Cr species with no Cr–Cr interactions, in O symmetry with an essentially trigonal environment. The second transition is longer lived (k = (2.09 ± 0.23) × 10⁻⁴ μs⁻¹) and is tentatively assigned to the symmetry forbidden ⁴T₂ → ⁴A₂ transition. The luminescence relaxation profiles of 1-Cr_b and 2-Cr_a (Fig. 4c) show similar bi-exponential decays, suggesting similar presence of multiple Cr(III) coordination environments.

Summary of spectroscopic results. In summary, there is no evidence from Raman or FTIR for incorporation of Cr(III) into the bulk structure of boehmite, consistent with ND and XRD results. In fact, XPS demonstrates that Cr(III) is concentrated towards the near-surface. Whereas ND indicates the presence of a Cr₂O₃ like environment in 1-Cr_b, electron (XPS) and emission spectroscopy indicate that Cr is likely in multiple coordination environments.

In the following section, electron microscopy provides direct spatial information on the distribution of Cr associated with both the rhombic and nano-lath boehmite morphologies.

Electron microscopy

Scanning electron micrographs (SEMs) of 1 show rhombic plates with a uniform size distribution (Fig. 5a). The high resolution transmission electron micrograph (TEM) of Fig. 5b shows the single crystal nature of the plates. The selected area electron diffraction (SAED) inset of Fig. 5b reveal the symmetric sharp diffraction spots that were indexed as orthorhombic γ-AlOOH with basal (010) and lateral (101) faces strongly expressed. The morphologies of 1-Cr_a and 1-Cr_b are identical to 1 (Fig. 5a) suggesting that Cr association with boehmite in the 1–10 wt% range does not affect particle morphology. Scanning transmission electron micrographs (STEMs) under high angle annular dark field (HAADF) mode and electron energy loss spectroscopy (EELS) analysis show ∼10 nm islets of Cr on the surface, Fig. 5c–f. Conversely, Cr was not detected in the bulk (detection limit ∼ 1 wt%). This preferential surface enrichment is also verified by focused ion beam (FIB) sections that expose the interior of the boehmite crystals, as shown in the energy-dispersive X-ray spectroscopy (EDX) mapping of Fig. 5g.

Fig. 5 Microanalyses of boehmite 1 and Cr-doped boehmite 1-Cr_a: representative (a) SEM, (b) TEM with SAED inset, (c) EELS mapping, (d) 2^nd derivative EELS spectrum of (c), (e) STEM HAADF of (c), (f) EELS spectra of (c)/(e), (g) STEM EDX mapping of FIB cross section.

Representative TEM of 2 and 2-Cr_a both show lath-like morphology, Fig. 6a. The selected area electron diffraction (SAED) inset of Fig. 6a reveal diffraction indexing as orthorhombic γ-AlOOH.

Fig. 6 Microanalyses of Cr doped boehmite 2-Cr_a: (a) TEM with SAED inset, (b) STEM HAADF, (c) EDX mapping of (b) (green spots represent Cr and purple spots represent Al).

A magnified section of the laths was examined with STEM under high angle annular dark field (HAADF) (Fig. 6b). The energy-dispersive X-ray spectroscopy (EDX) mapping (Fig. 6c) of the selected region shows the preferential enrichment of Cr on the surface. As in the case of 1-Cr_a, Cr was not detected in the bulk.

Implications from theory

In order to evaluate the stability of chromium in the boehmite structure, single Cr(III) ions or pairs of Cr(III) ions were substituted for lattice Al(III) ions in a boehmite supercell. Fig. 7a shows the 4 × 1 × 4 boehmite supercell where the substitution of 1 Cr represents 1.38 atomic weight%, and 2 Cr represents 2.76 atomic weight%. In Fig. 7b, calculated incorporation energies (E_inc-bulk) are plotted as a function of distance between two Cr(III) ions in the boehmite supercell normalized to the E_inc-bulk of 1 Cr(III). For comparison, the E_inc-bulk of a single Cr(III) substitution is shown at the minimum spacing between neighboring supercells (∼11 Å). In all cases, the E_inc-bulk are positive relative to gas-phase reference ions (eqn (S1) and Table S1; (ESI)†). Additional structural relaxation and hydration energy corrections would serve to help lower this E_inc-bulk; however, the strongly positive E_inc-bulk suggests that Cr(III) is unlikely to substitute for Al(III) in a bulk-like boehmite environment, or that a significant energy barrier must be overcome in order to stabilize Cr(III) in the structure.

Fig. 7 (a) Boehmite supercell (4 × 1 × 4) where pink spheres are Al, red spheres are O, and white spheres are H. Green sphere represents one Cr ion substituting for lattice one Al. Yellow spheres represent different locations of additional Cr ions added to evaluate E_inc-bulk for pairs of Cr ions. (b) Cr incorporation energy in bulk boehmite plotted as a function of distance between pairs of Cr(III) ions normalized to the energy of one Cr substituting for one Al (circles). E_inc-bulk for a single Cr(III) ion substituting for Al(III) ion is also shown (star). (c) Surface slab models for boehmite (010) illustrating Cr(III) substitution positions for Al(III) used to evaluate E_inc-slab as a function of Cr distance from the free surface of the slab. The (010) surface is parallel to the top and bottom of the slabs. Pink spheres are Al, red spheres are O, white spheres are H, and yellow spheres are Cr(III) substituting for Al(III).

Surface slab calculations were performed to test whether Cr(III) incorporation energies vary from the bulk to the (near-)surface. Vacuum-terminated surface energies were calculated for the commonly observed basal (010) and side, (100) and (001) surfaces of boehmite crystals (eqn (S2), ESI†). Due to challenges in stabilizing the commonly expressed (101) surfaces with vacuum terminations, those results are not included here. Single-point energy surface energies increase from the basal plane (010), 0.155 J m⁻², to the (001) and (100) surfaces, 1.689 J m⁻² and 3.226 J m⁻², respectively. These results suggest increasing surface reactivity from the basal planes to the sides under vacuum conditions. Next, chromium incorporation energies (E_inc-slab) were calculated as a function of Cr(III) distance from the free surface of the slab for all three terminations of boehmite. In order to avoid the formation of a dipole moment, Cr(III) were substituted for Al(III) in as symmetrically equivalent lattice locations as possible, moving from the center of the slab to the surface (Fig. 7c). While still positive, E_inc-slab decrease as the Cr ions move away from the bulk-like center of the slab to the (near-)surface slab environments suggesting that stability is gained by more surface-driven interactions of Cr with boehmite. For the basal (010) surface, this decrease in E_inc is only 0.004 eV going from the center of the slab to the top. For the side surfaces, the decrease in E_inc becomes more pronounced from the center to the top of the slab, decreasing by 0.603 eV for the (001) and 0.941 eV for the (100) surface slabs. In all cases, E_inc-slab for Cr in the center of the slab are consistent with E_inc-bulk (e.g., 2.922–2.926 eV), whereas (near-) surface E_inc-slab range from 1.981–2.922 eV.

In summary, the calculations are consistent with experiment and show that Cr(III) is incompatible with the boehmite structure at the ∼1% level, likely preferring a (near-)surface environment over a bulk-like environment. Given the large positive energies, it is reasonable to suggest that the solubility of Cr(III) in boehmite is appreciably lower than 1%. This conclusion is also consistent with the lack of Cr(III)–Al(III) phase formation during passivation of aluminum alloys that follow chromate/Al redox pathways.^26,27 From a size-based perspective, the crystal radius of high-spin Cr(III) in octahedral coordination is 11.8% larger than the radius of Al(III) (0.675 Å).²⁸ Generally, it is more difficult to substitute a large cation into a structure for a smaller cation;²⁹ however, according to Goldschmidt's rules of substitution, Cr(III) substitution for Al(III) could be plausible as their radii differ by less than 15%. However, if the ions form bonds of different ionic character as is the case for Al(III) versus Cr(III), (e.g., 3p versus 3d bonding orbitals, respectively), then substitution becomes more difficult. In addition, the fact that there is no naturally occurring Cr(III) polymorph equivalent to boehmite suggests that Cr(III) is not stable in the boehmite structure. Future work will focus on lowering the concentration of Cr(III) in the both the experiments and computations; however, there are practical limits to the size of the cell required to explore low Cr(III) concentrations.

Effect of exposure to caustic solutions: relevance to Hanford sludge waste processing

To investigate the possibility of Cr doping/adhesion effect on boehmite corrosion, samples 1 and 1-Cr_a were exposed to 2 M KOH solutions at 80 °C for 48 hours. The change in morphology due to corrosion was analyzed by SEM as shown in Fig. 8. The distinctive rhombohedral platelet shape is no longer visible for sample 1 as seen in Fig. 8a. The once sharp sides as seen in Fig. 5a, are now rounded with all particles on average significantly smaller in diameter than pre caustic treatment. In contrast 1-Cr_a shows little to no alteration under identical caustic conditions, Fig. 8b. Additionally the small particles of higher contrast in Fig. 8b, identified as Cr in STEM EDS mapping (ESI†), do not dissolve but instead remained adhered to the boehmite surface.

Fig. 8 SEM of 2 M KOH solution treatment at 80 °C for 48 h on sample (a) 1 and (b) 1-Cr_a.

Our preliminary observation is that Cr enrichment at rhombic boehmite surfaces (thus far there is no experimental or theoretical evidence for bulk incorporation of Cr) affects the reactivity of boehmite during caustic leaching. The inhibition or slowing the rate of corrosion by Cr additive has implications for the behavior of boehmite, and possibly other Al phases, in the Hanford tank waste.

For example, microanalyses and oxidative caustic-leaching (i.e., Cr(III) → Cr(VI)) of Cr-rich Hanford tank waste fractions indicate both a close association but poor correlation between Cr and Al during leaching (shown in Fig. 9). In particular, leaching the “group 6” sludge shows that a much greater fraction of Cr was mobilized initially compared to Al (Fig. 9d).¹⁵ After approximately 2 hours, Cr release plateaued but Al continued to be gradually released. One plausible interpretation that is bolstered by the present experimental and theoretical work is that the majority of Cr and Al are not forming a solid solution; rather Cr(III) and Al(III) are primarily forming separate phases. In other words, most of the Cr(III) is readily accessible for oxidation to highly soluble Cr(VI). Elsewhere in the tanks, Cr is a minor but not inconsequential constituent. However, further analysis is required to ascertain its relationship to boehmite in the general case. As established here, we predict that any Cr(III) that is associated with boehmite would be concentrated at the near-surface, not in the bulk. If so, oxidative leaching (preferably with a non-metallic oxidant) would indeed be an effective remedy for Cr removal as well as increasing the reactivity of boehmite to caustic leaching. There is also a possibility of some Al being present in the Cr-rich part, which would also support the trend observed in the oxidative leaching experiments.

Fig. 9 (a) TEM images of euhedral boehmite with interdispersed uranium oxide nanoclusters from a mixture of fully radioactive REDOX-type sludge from the Hanford site, (b) EDS analysis showing the potential association of Al and Cr in an amorphous phase from REDOX type wastes and (c) STEM image of the sludge material. (d) Kinetics of oxidative leaching of tank waste sludge from the REDOX waste. Orange circles = Cr(III), blue circles = Al(III) (Data adapted from Fiskum et al., 2008 (ref. 15)).

Conclusions

Irrespective of the starting materials, the synthetic method, morphology, or the stage of Cr introduction, Cr(III) is concentrated at the near-surface of boehmite and not in the bulk. The experimental results are consistent with quantum-mechanical calculations that yielded high positive incorporation energies for Cr(III) substituting for Al in the boehmite structure. We note that adsorption of Cr to surfaces, as opposed to bulk incorporation, could still strongly affect the energetics of boehmite, particularly nano-size variants due to the large contribution of surface energy to the overall free energy budget. Further, sorption of trace Cr(III) to surface defects could have a leveraged effect on dissolution kinetics. This has implications for interpreting the increased thermal stability of so called “Cr-doped” boehmite¹⁷ as well protocols designed to lower the Al load in HLW streams at the Hanford site.

Experimental section

Chemicals and materials

All chemicals were reagent grade or better. Gibbsite, Al(OH)₃(C-333) was obtained from Almatis, USA. Aluminium nitrate nonahydrate (Al(NO₃)₃·9H₂O; 98%) and chromium nitrate nonahydrate (Cr(NO₃)₃·9H₂O; 99%) were obtained from Sigma-Aldrich, USA. NaOH was obtained from Fischer Scientific, USA. DI water (18 MΩ) was used as a base for all solutions.

Synthesis of rhombic boehmite plates, 1. Gibbsite, Al(OH)₃ (0.023 mol) was added to a 30 ml 0.1 M NaOH aqueous solution at room temperature under stirring. Stirring was continued for a further 15 minutes to form a uniform suspension. The suspension was transferred to a Teflon-lined autoclave and kept at 200 ± 2 °C for 72 hours to yield a white solid. The obtained white solid was gravity-filtered, washed with excess water and methanol, air-dried and stored for further analyses.

Synthesis of 1 and 10 atom% Cr doped rhombic boehmite plates, 1-Cr_a and 1-Cr_b, respectively. The same procedure for synthesis 1 was followed except that the proportions were doubled and Cr(NO₃)₃·9H₂O (0.46 or 4.6 mmol) was added to the suspension prior to transfer to the Teflon-lined autoclave. The obtained solids had lilac (1% Cr) and suntanned (10% Cr) hues.

Synthesis of two-dimensional boehmite nano-laths, 2. The two-dimensional nano-laths were prepared by a modification of the method adapted by Chen et al.³⁰ 4.85 g of Al(NO₃)₃·9H₂O (13 mmol) and 0.96 g of urea (16 mmol) were sequentially dissolved in 70 ml DI water. The resultant mixture was stirred for a further 15 minutes and subsequently transferred to a Teflon-lined autoclave. The temperature of the resulting mixture was raised to 200 ± 2 °C and kept at this temperature for 24 hours to yield a white solid. The obtained solid was gravity-filtered, washed with excess water and methanol, air-dried and stored for further analyses.

Synthesis of 1 atom% Cr doped two-dimensional boehmite nano-laths, 2-Cr_a. Al(NO₃)₃·9H₂O (4.85 g, 13 mmol) and urea (16 mmol) were sequentially dissolved in 70 ml DI water at room temperature under stirring. Stirring was continued for a further 15 minutes to form a uniform suspension. To the resultant mixture, Cr(NO₃)₃·9H₂O (0.052 g, 0.13 mmol) was added and the resulting solution was allowed to stir for an additional 15 minutes to make the suspension uniform. The resultant mixture was transferred to a Teflon-lined autoclave and kept at 200 ± 2 °C for 24 hours to yield a lint colored solid. The obtained solid was gravity-filtered, washed with excess water and methanol, air-dried and stored for further analyses.

Solid phase characterization

The details of solid phase characterization techniques used in this work which include X-ray diffraction (XRD), time-of-flight (TOF) neutron diffraction, Fourier transform infra-red (FTIR), Raman, excitation and emission spectroscopies, X-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), scanning electron microscopy (SEM) and computational calculations are detailed in the ESI.†

Computational approach

To test the viability of Cr incorporation into the boehmite structure, atomic-scale models were developed representing a 4 × 1 × 4 supercell of boehmite (256 atoms) using the structure proposed by Corbato et al.²⁰ with Cmc2₁ space group symmetry. Additional symmetries were evaluated (ESI, Table S2†), with P2₁/b symmetry being applied to surface energy and slab incorporation energies. To complement Cr-doping levels in experimentally synthesized samples, one Al(III) center in the model was replaced by a Cr(III) center amounting to 1.38 atomic weight% of Cr(III) substitution. Models with pairs of Cr(III) replacing Al(III) (i.e., 2.76 atomic weight% of Cr(III)) were also evaluated to determine the effect of Cr-loading on incorporation energy. Incorporation energies for the charge-balanced substitution were calculated using the quantum-mechanical code, CRYSTAL14, at the unrestricted Hartree–Fock (UHF) level of theory.³¹ All-electron basis sets for Cr, Al, O, and H were chosen based on previous success modeling aluminum oxyhydroxides³² and chromium oxides.^33,34 Here, basis set performance was evaluated based on comparing the relaxed lattice parameters of complementary M₂O₃ structures (where M = Al(III) or Cr(III); ESI, Table S3†). Additionally, surface energy calculations were performed on commonly expressed, low-index surfaces of boehmite in order to determine their relative reactivity (eqn (S1), ESI†). Then, incorporation energies involving pairs of Cr(III) were performed on the most stable and more reactive terminations of boehmite to evaluate whether Cr(III) prefers bulk-like or (near-) surface environments when associating with this phase (eqn (S2), ESI†). Current calculations reflect single-point energies (i.e., no structural relaxation) at 0 K; however, additional structural relaxations and hydration energy corrections can be applied to link the calculations to more experimentally-relevant conditions.^35,36

Acknowledgements

This work was supported by the Laboratory Directed Research and Development (LDRD), Nuclear Processing Science Initiative (NPSI). Pacific Northwest National Laboratory (PNNL) is a multi-program national laboratory operated for the U.S. Department of Energy (DOE) by Battelle Memorial Institute under Contract DE-AC06-76RLO 1830. Part of this research was performed at EMSL, a national scientific user facility at PNNL managed by the Department of Energy's Office of Biological and Environmental Research. A portion of the research used resources at the Spallation Neutron Source (SNS), a DOE Office of Science User Facility operated by the Oak Ridge National Laboratory. Calculations were performed using PNNL Institutional Computing resources at Pacific Northwest National Laboratory.

References

1. Y. X. Zhang, X. B. Zhou, Z. L. Liu, B. Li, Q. C. Liu and X. H. Li, Monodispersed hierarchical gamma-AlOOH/Fe(OH)₍₃₎ micro/nanoflowers for efficient removal of heavy metal ions from water, RSC Adv., 2016, 6, 6695–6701.
2. T. Kim, H. B. Li, J. B. Lian, H. H. Jin, J. M. Ma, X. C. Duan, G. Yao and W. J. Zheng, Ionic liquid-assisted hydrothermal synthesis of gamma-Al₂O₃ hierarchical nanostructures, Cryst. Res. Technol., 2010, 45, 767–770.
3. M. Ireland, X. Wang, T. Radomirovic, P. Smith and F. Jones, Investigating the impact of anatase on the dissolution of boehmite, Hydrometallurgy, 2014, 147, 246–254.
4. K. Wefers, Nomenclature, preparation, and properties of aluminium oxide hydroxides, and trihydroxides, in Alumina Chemicals: Science and Technology Handbook, ed. L. D. Hart, The American Ceramic Society, Westerville, OH, 1990, pp. 13–22.
5. G. Martino, P. Courty, C. Marcilly, K. Kochloefl and J. H. Lunsford, Energy-Related Catalysis: Sections 3.1–3.7.5, VCH Verlag Gesellschaft, Weinheim, Germany, 1997.
6. M. Trueba and S. P. Trasatti, Gamma-Alumina as a support for catalysts: A review of fundamental aspects, Eur. J. Inorg. Chem., 2005, 3393–3403.
7. S. E. Fendorf, Surface-Reactions of Chromium in Soils and Waters, Geoderma, 1995, 67, 55–71.
8. G. E. Brown and G. A. Parks, Sorption of trace elements on mineral surfaces: Modern perspectives from spectroscopic studies, and comments on sorption in the marine environment, Int. Geol. Rev., 2001, 43, 963–1073.
9. R. E. Gephart, A short history of waste management at the Hanford Site, Phys. Chem. Earth, 2010, 35, 298–306.
10. J. G. Reynolds, G. A. Cooke, D. L. Herting and R. W. Warrant, Salt Mineralogy of Hanford High-Level Nuclear Waste Staged for Treatment, Ind. Eng. Chem. Res., 2013, 52, 9741–9751.
11. D. L. Herting, J. G. Reynolds and W. B. Barton, Conversion of Coarse Gibbsite Remaining in Hanford Nuclear Waste Tank Heels to Solid Sodium Aluminate [NaAl(OH)₍₄₎ center dot 1.5H₍₂₎O], Ind. Eng. Chem. Res., 2014, 53, 13833–13842.
12. J. G. Reynolds, G. A. Cooke, D. L. Herting and R. W. Warrant, Evidence for dawsonite in Hanford high-level nuclear waste tanks, J. Hazard. Mater., 2012, 209, 186–192.
13. J. G. Reynolds, J. K. McCoskey and D. L. Herting, Gibbsite Solubility in Hanford Nuclear Waste Approached from above and below Saturation, Ind. Eng. Chem. Res., 2016, 55, 5465–5473.
14. J. D. Vienna, P. Hrma, J. V. Crum and M. Mika, Liquidus temperature-composition model for multi-component glasses in the Fe, Cr, Ni, and Mn spinel primary phase field, J. Non-Cryst. Solids, 2001, 292, 1–24.
15. S. K. Fiskum, E. C. Buck, R. C. Daniel, K. E. Draper, M. K. Edwards, T. L. Hubler, L. K. Jagoda, E. D. Jenson, A. E. Kozelisky, G. J. Lumetta, P. J. MacFarlan, B. K. McNamara, R. A. Peterson, S. I. Sinkov, L. A. Snow and R. G. Swoboda, Characterization and Leach Testing for REDOX Sludge and S-Saltcake Actual Waste Sample Composites, Richland, 2008.
16. B. M. Rapko and J. D. Vienna, Selective Leaching of Chromium from Hanford Tank Sludge 241-U-108, Richland, WA, 2002.
17. J. Yang, R. L. Frost and Y. Yuan, Synthesis and characterization of chromium doped boehmite nanofibres, Thermochim. Acta, 2009, 483, 29–35.
18. R. L. Russell and R. A. Peterson, Boehmite Dissolution Model Based on Simulant Data, Ind. Eng. Chem. Res., 2010, 49, 4542–4545.
19. Powder Diffraction File Inorganic and Organic Data Book, ed. S. Kabekkodu, International Center for Diffraction Data, Newtown Square, PA, USA, 2015, vol. 60.
20. C. E. Corbato, R. T. Tettenhorst and G. G. Christoph, Structure Refinement of Deuterated Boehmite, Clays Clay Miner., 1985, 33, 71–75.
21. J. Zhang, S. Y. Wei, J. Lin, J. J. Luo, S. J. Liu, H. S. Song, E. Elawad, X. Ding, J. M. Gao, S. R. Qi and C. C. Tang, Template-free preparation of bunches of aligned boehmite nanowires, J. Phys. Chem. B, 2006, 110, 21680–21683.
22. H. D. Ruan, R. L. Frost and J. T. Kloprogge, Comparison of Raman spectra in characterizing gibbsite, bayerite, diaspore and boehmite, J. Raman Spectrosc., 2001, 32, 745–750.
23. J. T. Kloprogge, L. V. Duong, B. J. Wood and R. L. Frost, XPS study of the major minerals in bauxite: Gibbsite, bayerite and (pseudo-)boehmite, J. Colloid Interface Sci., 2006, 296, 572–576.
24. P. Marcus and J. M. Grimal, The Anodic-Dissolution and Passivation of Ni-Cr-Fe Alloys Studied by Esca, Corros. Sci., 1992, 33, 805–814.
25. S. Shoval, M. Gaft and G. Panczer, Luminescence of Cr³⁺ in natural and calcined diaspore, J. Therm. Anal. Calorim., 2003, 71, 699–706.
26. L. Xia and R. L. McCreery, Chemistry of a chromate conversion coating on aluminum alloy AA2024-T3 probed by vibrational spectroscopy, J. Electrochem. Soc., 1998, 145, 3083–3089.
27. J. D. Ramsey and R. L. McCreery, Raman microscopy of chromate interactions with corroding aluminum alloy 2024-T3, Corros. Sci., 2004, 46, 1729–1739.
28. R. D. Shannon, Revised Effective Ionic-Radii and Systematic Studies of Interatomic Distances in Halides and Chalcogenides, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751–767.
29. A. Navrotsky, Physics and Chemistry of Earth Materials, Cambridge University Press, New York, NY, 1994.
30. X. Y. Chen, H. S. Huh and S. W. Lee, Hydrothermal synthesis of boehmite (gamma-AlOOH) nanoplatelets and nanowires: pH-controlled morphologies, Nanotechnology, 2007, 18, 285608.
31. R. Dovesi, R. Orlando, A. Erba, C. M. Zicovich-Wilson, B. Civalleri, S. Casassa, L. Maschio, M. Ferrabone, M. De La Pierre, P. D'Arco, Y. Noel, M. Causa, M. Rerat and B. Kirtman, CRYSTAL14: A Program for the Ab Initio Investigation of Crystalline Solids, Int. J. Quantum Chem., 2014, 114, 1287–1317.
32. R. Demichelis, Y. Noel, P. Ugliengo, C. M. Zicovich-Wilson and R. Dovesi, Physico-Chemical Features of Aluminum Hydroxides As Modeled with the Hybrid B3LYP Functional and Localized Basis Functions, J. Phys. Chem. C, 2011, 115, 13107–13134.
33. M. Catti, G. Sandrone, G. Valerio and R. Dovesi, Electronic, magnetic and crystal structure of Cr₂O₃ by theoretical methods, J. Phys. Chem. Solids, 1996, 57, 1735–1741.
34. E. Ruiz, M. Llunell and P. Alemany, Calculation of exchange coupling constants in solid state transition metal compounds using localized atomic orbital basis sets, J. Solid State Chem., 2003, 176, 400–411.
35. F. N. Smith, C. D. Taylor, W. Um and A. A. Kruger, Technetium Incorporation into Goethite (alpha-FeOOH): An Atomic-Scale Investigation, Environ. Sci. Technol., 2015, 49, 13699–13707.
36. F. N. Smith, W. Um, C. D. Taylor, D. S. Kim, M. J. Schweiger and A. A. Kruger, Computational Investigation of Technetium(IV) Incorporation into Inverse Spinels: Magnetite (Fe₃O₄) and Trevorite (NiFe₂O₄), Environ. Sci. Technol., 2016, 50, 5216–5224.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20234a

---