New Data on the Isomorphism in Eudialyte-Group Minerals. 2. Crystal-Chemical Mechanisms of Blocky Isomorphism at the Key Sites

The review considers various complex mechanisms of isomorphism in the eudialyte-group minerals, involving both key positions of the heteropolyhedral framework and extra-framework components. In most cases, so-called blocky isomorphism is realized when one group of atoms and ions is replaced by another one, which is accompanied by a change in the valence state and/or coordination numbers of cations. The uniqueness of these minerals lies in the fact that they exhibit ability to blocky isomorphism at several sites of high-force-strength cations belonging to the framework and at numerous sites of extra-framework cations and anions.

. Arrangement of key sites in the eudialyte-type structures viewed along (210).
In addition to EGMs with the rhombohedral unit-cell parameters a ∼ 14.2 Å, c ∼ 30 Å, members of the eudialyte group with modular structures and doubled c parameter are known. Their unit cells contain two eudialyte-type modules which differ from each other by local situations around key sites [4].

Blocky Isomorphism at the M2 Site
The most complex blocky isomorphism is realized in the M2 micro-region situated between rings of octahedra, M16O24 [15]. This micro-region can be populated by cations having different radii, charges, and coordination (from flat square formed by edges of octahedra belonging to two Extra-framework cations (Na + , K + , Ca 2+ , Mn 2+ , Sr 2+ , Ba 2+ , Pb 2+ , Y 3+ , Ln 3+ , and H 3 O + ) and (in some samples) water molecules occupy five sites, N1-N5, which are typically split and can be partly vacant. Cations other than Na + show a tendency to concentrate at the N3 and N4 sites. Some of these cations (K + , Ca 2+ , Mn 2+ , Sr 2+ , Ce 3+ , and H 3 O + ) are species-defining ones in several representatives of the eudialyte group. The Ø anions (Ø = O, OH) coordinate the M2, M3, and M4 sites. Additional anions (Cl − , F − , OH − , S 2− , SO 4 2− , and CO 3 2− ) and water molecules occur at the X1 and X2 sites located on the three-fold axis. Blocky isomorphism is defined as the ability of groups of atoms or ions having different configurations to replace each other in crystal structures [14]. Such substitutions are known for a large number of alkaline zircono-and titanosilicates [9]. In EGMs this kind of isomorphism is realized at the key sites M2, M3, and M4, as well as at the N and X sites. The eudialyte group is the only group of minerals in which blocky isomorphism is realized at several sites containing high-force-strength cations. Below we will use the symbols N1-N5, M2, M3, M4, and X to denote corresponding cavities (i.e., micro-regions which can contain several closely spaced sites).
In addition to EGMs with the rhombohedral unit-cell parameters a~14.2 Å, c~30 Å, members of the eudialyte group with modular structures and doubled c parameter are known. Their unit cells contain two eudialyte-type modules which differ from each other by local situations around key sites [4].

Blocky Isomorphism at the M2 Site
The most complex blocky isomorphism is realized in the M2 micro-region situated between rings of octahedra, M1 6 O 24 [15]. This micro-region can be populated by cations having different radii, charges, and coordination (from flat square formed by edges of octahedra belonging to two neighboring M1 6 O 24 rings to a 7-or 8-fold polyhedron): IV Fe 2+ , V Fe 2+ , V Fe 3+ , VI Fe 3+ , V Mn 2+ , VI Mn 2+ , IV Zr, IV Ta, IV Na, V Na, VI Na, VII Na, VI K, VII K, and VIII K. The M2-cations occur in the plane of the six-membered rings which is perpendicular to the threefold axis ( Figure 2). The coordination polyhedra M2O 5 and M2O 7 can have different orientations with respect to the plane of the square. Some examples of M2-centered polyhedra are shown in Figure 3.

Blocky Isomorphism at the M3 and M4 Sites
The micro-regions M3 and M4 at the centers of two nonequivalent Si 9 O 27 rings can be vacant, but usually they contain tetrahedra (SiO 4 , rarely AlO 4 ) or octahedra (typically, NbO 6 ; rarely, TiO 6 , WO 6 , MnO 6 , and NaO 6 ) [1] ( Figure 5). The positions of M3 and M4 cations are located on the threefold axis, and their charge can vary from +1 (Na) to +6 (W). When SiO 4 tetrahedron occurs in the M3 or M4 micro-region, corresponding Si 9 O 27 ring transforms into a 10-membered disc consisting of three 5-membered rings ( Figure 5b). The VII M2-centered sevenfold polyhedra are based on the flat square and include O atoms of the framework and large X anions. Positions of large cations (K + , Sr 2+ , or Ce 3+ ) occurring in the M2 micro-region, are located at large distances from the plane of the O4 square. In these cases, the coordination polyhedron is supplemented by two inner O atoms of the Si9O27 ring and one or two (statistically) species anions (Cl, F, or H2O) situated at the X site on the threefold axis. As a result, a 7or 8-fold polyhedron is formed. The coordination number of Na + occurring in the M2 micro-region can vary from 4 to 7. The VII Na-centered polyhedron dominates in the M2 micro-region of some samples [28,44]. As a subordinate component, M2 ( VII Na) occurs in the structures of intermediate members of the manganoeudialyte-ilyukhinite [45] and eudialyte-sergevanite [46] solid-solution series, with the Na-O distances of 2.23(4)-2.96(3) and 2.33(1)-3.01(1) Å, respectively. Rarely, the VII M2 polyhedron is centered by Mn or Zr.

Blocky Isomorphism at the M3 and M4 Sites
The micro-regions M3 and M4 at the centers of two nonequivalent Si9O27 rings can be vacant, but usually they contain tetrahedra (SiO4, rarely AlO4) or octahedra (typically, NbO6; rarely, TiO6, WO6, MnO6, and NaО6) [1] ( Figure 5). The positions of M3 and M4 cations are located on the threefold axis, and their charge can vary from +1 (Na) to +6 (W). When SiO4 tetrahedron occurs in the M3 or M4 micro-region, corresponding Si9O27 ring transforms into a 10-membered disc consisting of three 5-membered rings (Figure 5b). Three O atoms which form the base of each Si-centered tetrahedron occurring in the M3 or M4 micro-region are common with O atoms of the Si9O27 ring. The fourth O atom of this tetrahedron belongs to OH group. The M3,M4 Si atoms can occur either above or below the triangular base of the Three O atoms which form the base of each Si-centered tetrahedron occurring in the M3 or M4 micro-region are common with O atoms of the Si 9 O 27 ring. The fourth O atom of this tetrahedron belongs to OH group. The M3,M4 Si atoms can occur either above or below the triangular base of the tetrahedron, and OH groups belonging to the tetrahedra can be located on both sides of the Si 9 O 27 ring in such a way that apical vertices of the neighboring tetrahedra are oriented in opposite directions. Different options involving M3 and M4 tetrahedra are shown in Figure 6. It is to be noted that the M3a and M3b tetrahedra, as well as M4a and M4b tetrahedra cannot exist in corresponding micro-regions simultaneously because of short M3a Si-M3b Si and M4a Si-M4b Si distances. The option (b) corresponds to a centrosymmetric structure, and the options (d) and (e) correspond to non-centrosymmetric ones.
The option (c) is most common, but corresponding local situation cannot be realized due to unacceptably short OH-OH distance of 1.4-1.5 Å between the vertices of the tetrahedra. However, this option can be realized statistically.
An example of the option (a) is the crystal structure of a high-silicon EMG from the Kovdor massif, Kola Peninsula, described as "low-iron eudialyte 2" [50]. In this sample, the M3 and M4 sites occupied by Si are split into pairs of partially occupied sub-sites located at short distances from each other: M3a-M3b = 0.92(6) Å and M4b-M4c = 1.04(1) Å. As noted above, a local situation when both M3b and M4a sites are occupied is impossible due to a short distance (of 1.41 Å) between OH groups belonging to corresponding tetrahedra.
An example of the option (a) is the crystal structure of a high-silicon EMG from the Kovdor massif, Kola Peninsula, described as "low-iron eudialyte 2" [50]. In this sample, the M3 and M4 sites occupied by Si are split into pairs of partially occupied sub-sites located at short distances from each other: M3а-M3b = 0.92(6) Å and M4b-M4c = 1.04(1) Å. As noted above, a local situation when both M3b and M4a sites are occupied is impossible due to a short distance (of 1.41 Å) between OH groups belonging to corresponding tetrahedra. Our investigations carried out on a large selection of EGMs [5,51] (Table 2) have shown that the total amount of Si atoms at the M3 and M4 sites (∑T) varies in wide ranges, from 0.88 to almost 2 atoms per formula unit (apfu) (Z = 3). The ratio of the amounts of differently oriented tetrahedra (∑Ta/∑Tb) varies from 0 to 2.31; only in two structures this ratio is equal to 1 which corresponds to the existence of center of symmetry. Both centrosymmetric structures are characterized by rather low ∑T values of 1.12 and 1.39 apfu, and only one of them corresponds to the option (a). Our investigations carried out on a large selection of EGMs [5,51] (Table 2) have shown that the total amount of Si atoms at the M3 and M4 sites ( T) varies in wide ranges, from 0.88 to almost 2 atoms per formula unit (apfu) (Z = 3). The ratio of the amounts of differently oriented tetrahedra ( T a / T b ) varies from 0 to 2.31; only in two structures this ratio is equal to 1 which corresponds to the existence of center of symmetry. Both centrosymmetric structures are characterized by rather low T values of 1.12 and 1.39 apfu, and only one of them corresponds to the option (a). Table 2. Population of tetrahedral sites in the M3 and M4 micro-regions, total amounts of the IV M3 and IV M4 components ( T, Z = 3) and atomic ratios of the contents of ( IV M3a, IV M4a) and ( IV M3b, IV M4b) cations ( T a / T b ). The samples are listed in the order of lowering of the T value.

No.
Mineral   [19] In Si-deficient EGMs, the M3 and M4 micro-regions contain vacancies (as e.g., in ikranite [57,62,63] and mogovidite [38,64]) and/or octahedra formed by three O atoms common with the Si 9 O 27 rings and three OH groups. The M3 and M4 cations with octahedral coordination can occur in one of three sites located at short distances from the IV M3 and IV M4 sites which are usually occupied by Si. Consequently, the IV M3 and VI M3 sites, as well as the IV M4 and VI M4, cannot be occupied simultaneously in the same micro-region. In most cases, blocky isomorphism involving atoms with tetrahedral and octahedral coordination is realized in the M3 and M4 micro-regions. However, there are cases when these microregions contain only one type of polyhedra, namely, only tetrahedra (for example, in davinciite [52,53]), or only octahedra. The latter case is realized in the centrosymmetric high zirconium eudialyte described by Giuseppetti [65], in which the M3 and M4 sites have been refined with the same occupancy, 0.65 VI Zr 0.35 . However, the possibility of the occurrence of Zr in the M3 and M4 micro-regions was not confirmed by subsequent investigations. In most cases, octahedral sites in these micro-regions are occupied by Nb. In three EGMs (khomyakovite and manganokhomyakovite [66], and johnsenite-(Ce) [          Unlike Si-centered tetrahedra, the NbO 3 (OH) 3 octahedra located in the M3 and M4 micro-regions can be directed towards the N5 cavity between the neighboring Si 9 O 27 rings only in the case when the N5 site is vacant. Otherwise this would result in unrealistic short distances between OH groups and N5 cations.
The option when the octahedra are oriented both outward and inward of the N5 cavity is extremely rare. This situation takes place in a low-silicon sample with high contents of niobium and titanium and a low content of zirconium, found in the Kovdor phlogopite deposit, Kola Peninsula ( Figure 10). This centrosymmetric member of the eudialyte group was formed as the result of partial leaching of sodium from an earlier EGM [68]. In the micro-regions M3 and M4 of this mineral, there are two octahedral sites located at a distance of 2.19 Å from each other and populated by titanium with the occupancy factors of 50 and 30%, respectively. A TiO 6 octahedron with a population of 50% is directed inside the N5 cavity, while the less populated one faces in the opposite direction from the 9-membered ring. At a distance of 0.29 Å from the Ti site, there is a site occupied by Si with the population of 20%. Thus, all positions in the M3 and M4 micro-regions are populated statistically. leaching of sodium from an earlier EGM [68]. In the micro-regions M3 and M4 of this mineral, there are two octahedral sites located at a distance of 2.19 Å from each other and populated by titanium with the occupancy factors of 50 and 30%, respectively. A TiO6 octahedron with a population of 50% is directed inside the N5 cavity, while the less populated one faces in the opposite direction from the 9-membered ring. At a distance of 0.29 Å from the Ti site, there is a site occupied by Si with the population of 20%. Thus, all positions in the M3 and M4 micro-regions are populated statistically.

Blocky Isomorphism at N Sites
At the N1-N5 extra-framework sites occurring in large cavities of the structures of EGMs, positional isomorphism of a statistical nature is most often observed. Below we will use the symbols N1-N5 to denote corresponding cavities (i.e., micro-regions containing several closely spaced extra-framework sites). In most cases, these positions are split and occupied by sodium; the set of closely spaced Na sites is typically coordinated by anions belonging to the same polyhedron. However, if significant amounts of other large cations enter these sites, ordering of different elements among closely spaced sites may result in changes in their anionic environment.
Most frequently, extra-framework cations other than Na occur at the N3 and N4 sites (Table 3). For example, in the structure of davinciite [52], the N3 and N4 cavities contain pairs of partially populated sites with the distances N3a-N3b = 0.65(1) Å and N4a-N4b = 0.42(1) Å. The N3a and N3b sites are partially occupied by K and (K + Na) and have 7-and 6-fold coordination, respectively. The 6-fold polyhedron is formed solely by O atoms of the framework, whereas the 7-fold polyhedron contains one H2O molecule. The N4a and N4b sites are partially occupied by (Na + Sr) and Sr and have 8-and 7-fold coordination, respectively. The larger Sr-centered 8-fold polyhedron involves two Cl atoms, and smaller one contains only one Cl atom.

Blocky Isomorphism at N Sites
At the N1-N5 extra-framework sites occurring in large cavities of the structures of EGMs, positional isomorphism of a statistical nature is most often observed. Below we will use the symbols N1-N5 to denote corresponding cavities (i.e., micro-regions containing several closely spaced extra-framework sites). In most cases, these positions are split and occupied by sodium; the set of closely spaced Na sites is typically coordinated by anions belonging to the same polyhedron. However, if significant amounts of other large cations enter these sites, ordering of different elements among closely spaced sites may result in changes in their anionic environment.
Most frequently, extra-framework cations other than Na occur at the N3 and N4 sites (Table 3). For example, in the structure of davinciite [52], the N3 and N4 cavities contain pairs of partially populated sites with the distances N3a-N3b = 0.65(1) Å and N4a-N4b = 0.42(1) Å. The N3a and N3b sites are partially occupied by K and (K + Na) and have 7-and 6-fold coordination, respectively. The 6-fold polyhedron is formed solely by O atoms of the framework, whereas the 7-fold polyhedron contains one H 2 O molecule. The N4a and N4b sites are partially occupied by (Na + Sr) and Sr and have 8-and 7-fold coordination, respectively. The larger Sr-centered 8-fold polyhedron involves two Cl atoms, and smaller one contains only one Cl atom.
A remarkable example of blocky isomorphism at the N sites is presented in a Fe-and Na-deficient EGM [41,50,69]. In this sample, sodium deficiency of almost 50% compared to eudialyte s.s. is mainly compensated by hydronium groups H 3 O + , as well as subordinate K + , Sr 2+ , and Ce 3+ . These cations occur in the N3 cavity and are statistically distributed over four sites, located at distances of 0.32(1) to 0.95(1) Å from each other. The coordination numbers and cation-anion distances (Å) for these sites are 10 and 2.91 for H 3 O, 10 and 2.87 for K, 9 and 2.73 for Ce, and 7 and 2.61 for Sr.
Large N4-cations form triads of edge-sharing polyhedra. The coordination number of the N4 site in kentbrooksite, taseqite, and georgbarsanovite is equal to 10 ( Figure 11) and is characterized by the cation-anion distances from 2.471 to 2.943 Å. The N4 site in andrianovite has 8-fold coordination and the cation-anion distances from 2.557 to 2.980 Å. Mixed population is typical for the N4 site. For example, the composition of the N4 micro-region in georgbarsanovite and andrianovite is (Mn 1.07 Sr 0.72 Ce 0.47 Ca 0.33 K 0.25 Y 0.16 ) 3 and (K 1.45 Sr 1.05 Ce 0.5 ) 3 , respectively.
In the majority of eudialyte-related minerals, the crystal chemical formula (Z = 3) contains 6 Ca atoms, which enter into the composition of the six-membered ring of octahedra, and the N3 and N4 sites are predominantly occupied by Na (Figure 12a). However, in specific environments, EGMs of a different stoichiometry are formed. In particular, there are three representatives of this group with a high content of calcium of 9 to 10 Ca apfu: feklichevite, golyshevite, and mogovidite (Table 1).
An example of EGM with an unusually high Ca content is feklichevite, Na11Ca9Fe2Zr3NbSi25O73(OH,H2O,O,Cl)5 [37]. In this mineral, Ca completely occupies the M1 site in the ring of octahedra and dominates at the neighboring N3 site. A cluster formed by three Ca-centered polyhedra sharing common vertices occurs at the threefold axis completed by six M1 octahedra (Figure 12b).
In the majority of eudialyte-related minerals, the crystal chemical formula (Z = 3) contains 6 Ca atoms, which enter into the composition of the six-membered ring of octahedra, and the N3 and N4 sites are predominantly occupied by Na (Figure 12a). However, in specific environments, EGMs of a different stoichiometry are formed. In particular, there are three representatives of this group with a high content of calcium of 9 to 10 Ca apfu: feklichevite, golyshevite, and mogovidite (Table 1).
An example of EGM with an unusually high Ca content is feklichevite, Na 11 Ca 9 Fe 2 Zr 3 NbSi 25 O 73 (OH,H 2 O,O,Cl) 5 [37]. In this mineral, Ca completely occupies the M1 site in the ring of octahedra and dominates at the neighboring N3 site. A cluster formed by three Ca-centered polyhedra sharing common vertices occurs at the threefold axis completed by six M1 octahedra (Figure 12b).
Golyshevite and mogovidite are the most Ca-rich EGMs (Table 3). In these minerals, Ca occupies the M1 site in the ring of octahedra and dominates over Na at N(4) (in golyshevite [39], Figure 12c) or at both N(3) and N(4) (in mogovidite [38]). In the latter case, a double layer of Ca-centered polyhedra is formed (Figure 12d).
The N3and N4-centered polyhedra share common edges with the M1O 6 octahedra. As a result, the substitution of Na by bi-or trivalent cations at the N3 and N4 sites results in changes of local situations in the M2 micro-region and formation of rigid layers formed by the M1-, M2-, N3-, and N4-centered polyhedra. Acentric character of the georgbarsanovite crystal structure is especially obvious because of the presence of bi-and trivalent cations at the N4 site. As a result, this mineral demonstrates a pronounced piezo-effect.
The following local situations around the M2 micro-region are possible in Na-deficient EGMs:  Table 3. Dominant components at the N3 and N4 sites of Na-depleted EGMs.

Blocky Isomorphism at the X Sites
In the structures of EGMs, the extra-framework X1 and X2 sites located on the threefold axis can be occupied by different anions (Cl − , F − , S 2− , SO 4 2− , and CO 3 2− ) and H 2 O molecules which coordinate large N and (in some samples) M2 cations. The most common X-components are Cl − (e.g., in eudialyte s.s.) and H 2 O (e.g., in feklichevite). Carbonate anion is a significant component in five EGMs: golyshevite, mogovidite, carbokentbrooksite, johnsenite-(Ce), and zirsilite. The only EGM with species-defining F − is kentbrooksite. S-bearing anions can occur at the X sites in subordinate amounts.
In zirsilite-(Ce) and carbokentbrooksite, the X1 site is split into two sub-sites, one of which is partly occupied by CO 3 2− and another one contains Cl − . The total population of the X1 site is C 0. 43 Cl 0.3 in zirsilite-(Ce) and C 0. 58 Cl 0.27 in carbokentbrooksite. The X2 site in these minerals is occupied by H 2 O. In golyshevite, the X1 site is split into two sub-sites located at the distance of 0.55 Å from each other and occupied by CO 3 2− (with the total occupancy close to 100%), and the X2 site is occupied by H 2 O and subordinate Cl − . Both carbonate groups are flat triangles whose O atoms coordinate cations of the N4 micro-region ( Figure 9). In most S-bearing EGMs, sulfur occurs in the sulfate form. In these minerals, isolated SO 4 tetrahedra occur at the sites located on the threefold axis (see Figures 10 and 13). A potentially new EGM containing sulfur in the sulfide form has been discovered recently in the Lovozero alkaline massif, Kola Peninsula [72]. Hypothetically, this mineral crystallized at a low activity of oxygen. Its end-member formula is (Na 14 Sr)Ca 6 (Mn 2 Na)Zr 3 [Si 26 O 72 ](OH) 3 S 2− ·2H 2 O. In the sulfide EGM, the X1 and X2 sites are split into pairs of closely spaced sub-sites with the following occupancies:

Ordered Blocky Isomorphism in Modular EGMs
Heteropolyhedral frameworks of EGMs consist of layers containing Si 3 O 9 and Si 9 O 27 rings of tetrahedra (T-layer), VI M1 6 O 24 rings of octahedra (M-layer), and isolated ZO 6 octahedra (Z-layer) alternating in the order ( . . . TMTZ . . . ). Most EGMs have 12-layer structures, with the 2D repeat unit (TMTZ) 3 , unit-cell parameters a~14 Å, c~30 Å, and space groups R3m, R3m, or R3. However, in several members of the eudialyte group (so-called "megaeudialytes"), additional cation ordering takes place which results in the formation of the doubled 2D repeat units [(TMTZ)(T*M*T*Z*)] 3 with the ordered 1:1 alternation of two nonequivalent TMTZ modules and doubling of the c parameter (c~60-61 Å). Megaeudialytes are characterized by the same space groups as common 12-layer EGMs. Unlike crystalline compounds with superstructures, in megaeudialytes R-translation remains, and the unit-cell doubling is due to blocky isomorphism in the 30 Å module.
Megaeudialytes are known only in the Lovozero and Khibiny alkaline massifs, Kola Peninsula [1,4,[73][74][75][76][77][78][79]. Four of them (alluaivite, dualite, labyrinthite, and rastsvetaevite) have been approved by the Commission on New Minerals, Nomenclature, and Classification of the IMA as valid mineral species (see Figures 14 and 15).  Unlike zirconium, titanium rarely forms isolated octahedra in structures of silicate minerals [80]. The main cause of this distinction between the behavior of these elements is a different degree of distortion of Zr-and Ti-centered octahedra. As a result, the isomorphism between Ti and Zr is usually significantly restricted. Minerals of the eudialyte group are one of exceptions to this rule. In EGMs, titanium can occur at the M2, M3, and M4 sites (Table 4). Unlike zirconium, titanium rarely forms isolated octahedra in structures of silicate minerals [80]. The main cause of this distinction between the behavior of these elements is a different degree of distortion of Zr-and Ti-centered octahedra. As a result, the isomorphism between Ti and Zr is usually significantly restricted. Minerals of the eudialyte group are one of exceptions to this rule. In EGMs, titanium can occur at the M2, M3, and M4 sites (Table 4).  [76] Note: Orientation of Si-centered tetrahedra at the M3 and M4 sites of alluaivite is shown with arrows.
Alluaivite (Sample 1 in Table 4; Figure 15) [81] is the first megaeudialyte whose crystal structure was solved. In this mineral zirconium is absent, and the Z and Z* sites are occupied by Ti and subordinate Nb. The substitution of Zr 4+ by the smaller cation Ti 4+ results in the lowering of unit-cell parameters (by 0.1-0.2 Å for a and 0.1-0.3 Å for c). Both the M2 and M2* micro-regions (in the modules I and II, respectively) are occupied by Na. All Na in the M1 micro-region has fourfold flat-square coordination. In the M2 micro-region, Na is distributed among three sites: one site having flat-square coordination and two sites with 7-fold coordination located on both sides from the square. Another cause of the unit-cell doubling is blocky isomorphism in the M3 and M4 micro-regions; in each module, both additional SiO 4 tetrahedra have the same orientation, but their orientation is different in different modules.
In the structure of dualite (Sample 3 in Table 4; see Figure 14a) [43], alternation of Zr and Ti in the modules I and II takes place. The M3 + M4 and M3* + M4* micro-regions of dualite are dominated by IV Mn and Si, respectively. There are also significant differences in the populations of the V Na-dominant M2 and M2* sites of dualite.
The unit-cell doubling of rastsvetaevite (Sample 4 in Table 4; see Figure 14b) [78] is mainly due to the blocky isomorphism in the M2 and M2* micro-regions: in M2, Na has fourfold flat-square coordination and is surrounded by K atoms having sevenfold coordination; the micro-region M2* is predominantly occupied by Fe. There are also differences between the modules I and II of rastsvetaevite in the population of extra-framework sites. In particular, in the module I the N3 site is occupied by K, whereas in the module II all N sites are Na-dominant.
In most cases, the formation of megaeudialytes is regulated by the ordering of the M2-cations. For example, in hydrated rastsvetaevite (Sample 10 in Table 4; see Figure 14c) [75], different population of the M2 and M2* micro-regions (by IV Na 2.7 V Fe 0.3 and V Fe 2.0 V Mn 0.8 , respectively) is the main cause of the unit-cell doubling.

Combinations of Blocky Isomorphism at the M1-M4 Sites
The population of each of the M1-M4 micro-regions does not occur independently on the local situations at neighboring sites. Very often, M2-centered polyhedra share common edges with the M1O 6 octahedra and common vertices with the M3O 6 and M4O 6 octahedra. As a result, only specific kinds of clusters (i.e., combinations of local situations around neighboring M1-M4 cations) can be realized with their statistical alteration in different unit cells. The examples of different clusters which can be involved in the combined blocky isomorphism in EGMs with the symmetry R3m (1-4) and R3 (5-8) are The M1O 6 octahedra can be occupied by different elements including Ca, Mn, Fe, Na, REE, Sr, and Y. For example, in one module of dualite, all the listed elements, except Y statistically substitute each other [43].

Combinations of Blocky Isomorphism at the M1-M4 Sites
The population of each of the M1-M4 micro-regions does not occur independently on the local situations at neighboring sites. Very often, M2-centered polyhedra share common edges with the M1O6 octahedra and common vertices with the M3O6 and M4O6 octahedra. As a result, only specific kinds of clusters (i.e., combinations of local situations around neighboring M1-M4 cations) can be realized with their statistical alteration in different unit cells. The examples of different clusters which can be involved in the combined blocky isomorphism in EGMs with the symmetry R3m (1-4) and R3 (5)(6)(7)(8)  The M1O6 octahedra can be occupied by different elements including Ca, Mn, Fe, Na, REE, Sr, and Y. For example, in one module of dualite, all the listed elements, except Y statistically substitute each other [43].   In kentbrooksite, barsanovite and some related EGMs, the M2 Mn 2+ O5 and M2 Fe 2+ O5 polyhedra are combined with the M3,M4 Nb 5+ O6 octahedra. In W-dominant EGMs related to kentbrooksite (i.e., khomyakovite, manganokhomyakovite, and johnsenite-(Ce)), the M2 Mn 2+ O5 and M2 Fe 2+ O5 pyramids are combined with the M3,M4 W 6+ O6 octahedra. In these cases, local charge balance is regulated by the isomorphism between O and OH at the common vertex of the square pyramid and octahedron.
A high degree of separation of different elements in the M2 micro-region is realized in a Fe-deficient sample from the Oleniy Ruchey deposit, Khibiny massif, Kola Peninsula [10]. Its crystal-chemical formula is (Z = 3): A more complex situation takes place in Ca-depleted samples, in which ordering in the M2 micro-region is accompanied by the ordering in the M1 sites of the 6-membered ring of octahedra. In this case, the M1 site is transformed into two alternating sites, M1.1 and M1.2 (Figure 18), which are occupied differently (respectively, by Ca and Mn in oneillite and sergevanite, by Ca and Fe in raslakite, and by Mn and Na in voronkovite). Differently occupied octahedra have different sizes (Table 5)   In kentbrooksite, barsanovite and some related EGMs, the M2 Mn 2+ O 5 and M2 Fe 2+ O 5 polyhedra are combined with the M3,M4 Nb 5+ O 6 octahedra. In W-dominant EGMs related to kentbrooksite (i.e., khomyakovite, manganokhomyakovite, and johnsenite-(Ce)), the M2 Mn 2+ O 5 and M2 Fe 2+ O 5 pyramids are combined with the M3,M4 W 6+ O 6 octahedra. In these cases, local charge balance is regulated by the isomorphism between O and OH at the common vertex of the square pyramid and octahedron.
A high degree of separation of different elements in the M2 micro-region is realized in a Fe-deficient sample from the Oleniy Ruchey deposit, Khibiny massif, Kola Peninsula [10]. Its crystal-chemical formula is (Z = 3): A more complex situation takes place in Ca-depleted samples, in which ordering in the M2 micro-region is accompanied by the ordering in the M1 sites of the 6-membered ring of octahedra. In this case, the M1 site is transformed into two alternating sites, M1.1 and M1.2 (Figure 18), which are occupied differently (respectively, by Ca and Mn in oneillite and sergevanite, by Ca and Fe in raslakite, and by Mn and Na in voronkovite). Differently occupied octahedra have different sizes (Table 5) which results in variations of configurations of M2-polyhedra.

Conclusions
The uniqueness of the eudialyte-group minerals lies in the combination of complexity and variability of their crystalline structures and various schemes of homovalent, heterovalent, and blocky isomorphism, in which at least 26 different elements are involved. From the crystal chemical point of view, the most interesting are the complex mechanisms of blocky isomorphism, when, in a

Conclusions
The uniqueness of the eudialyte-group minerals lies in the combination of complexity and variability of their crystalline structures and various schemes of homovalent, heterovalent, and blocky isomorphism, in which at least 26 different elements are involved. From the crystal chemical point of view, the most interesting are the complex mechanisms of blocky isomorphism, when, in a given