Eur. J. Mineral.

Dedicated to the memory of

2003, 15, 419–426 Luciano Ungaretti

A new occurrence of katoite and re-examination of the hydrogrossular group

ORAZIO FERRO1, ERMANNO GALLI1, GÁBOR PAPP2, SIMONA QUARTIERI3, SÁNDOR SZAKÁLL4
and GIOVANNA VEZZALINI1,*

1
Dipartimento di Scienze della Terra, via S. Eufemia 19, I-41100 Modena, Italy
* Corresponding author, e-mail: giovanna@unimo.it
2 Dept. of Mineralogy and Petrology, Hungarian Natural History Museum, Ludovika tér 2, H-1083 Budapest, Hungary
3 Dipartimento di Scienze della Terra, Salita Sperone 31, I-98166 Messina S. Agata, Italy
4 Dept. of Mineralogy, Herman Ottó Museum, Kossuth u. 13, H-3525 Miskolc, Hungary

Abstract: A new occurrence of katoite (Ca2.95Fe0.03Al2.03(SiO4)1.12(OH)7.51; a = 12.286) found near Dunabogdány (Hungary), along
with its structural refinement is reported here. The crystal-chemical data, the physical properties, and the structure refinement of this
new member of the Ca3Al2(SiO4)3–Ca3Al2(O4H4)3 series indicate the presence of about 37% grossular. The positional disorder of the
oxygen atom, a peculiar feature of the hydrogarnet structure, is described here in terms of both anisotropy along the d-O vector, and
presence of two distinct oxygen sites (unsplit- and split-O models, respectively). The hydrogen atom was localised for both models
in the difference Fourier map. The structural features of the katoite from Dunabogdány are compared with those of other
hydrogrossulars and with the two end-members “anhydrous” grossular, and Si-free katoite. The results of the unsplit-O model
refinement confirm the increase of the tetrahedron volume with the substitution of Si by 4H+. The geometry of the coordination
polyhedra in the two configurations resulting from the split-O model are also discussed. This second model better describes from the
crystallochemical point of view the substitution of Si4+ with 4H+ in the tetrahedra. However, the tetrahedra of the two dimensions are
randomly distributed in the unit cell and hence domains of the two end-member configurations are not expected in katoite structure.
This conclusion is also strengthened by the TEM study of katoite from Dunabogdany.**
Key-words: katoite, hydrogrossular, garnet, crystal structure, chemical analysis.

Introduction
The hydrogrossular group can be described as a solid solu-
tion between grossular, (Gr) Ca3Al2(SiO4)3 (s.g. Ia-3d, a =
11.845(1)Å; Novak & Gibbs, 1971) and Si-free hydrogros-
sular Ca3Al2(O4H4)3 (a = 12.5695(1) Å; Lager et al., 1987a).
Following the nomenclature proposed by Passaglia & Ri-
naldi (1984), the minerals of this series with a Gr component
& 50% are called hibschites, whereas those with Gr < 50%
are called katoites. The same authors state that the distinc-
tive properties of the katoite members are: refractive index
n < 1.67, unit cell edge larger than 12.21 Å, presence of the
(220) line in the X-ray powder pattern. Structural and crys-
tal-chemical data on natural and synthetic hydrated mem-
bers of this series are available for: hibschites (Basso et al.,
1983; Lager et al., 1987b; Armbruster & Lager, 1989), kato-
ite (Passaglia & Rinaldi, 1984; Sacerdoti & Passaglia, 1985;
Armbruster & Lager, 1989) and synthetic Si-free hydro-
grossular (Cohen-Addad et al., 1967; Lager et al., 1987a).
Minor OH– substitution is also reported for garnets in the
grossular-andradite-almandine-spessartine system (Hsu,
1980; Basso et al., 1981; Basso et al., 1984a, b; Lager et al.,
1989; Armbruster, 1995).
In the hydrogrossular where H substitutes for silicon (in
d position, site symmetry –4) each oxygen around the tetra-
hedral void is bonded to one H. As a consequence the d-O
distance lengthens from 1.641 Å in grossular (Novak &
Gibbs, 1971) to 1.950 Å in the silicon-free member (Lager
et al., 1987a). Because of the random arrangement of SiO4
and (O4H4) tetrahedra in the structure, the space group sym-
metry is unaffected by the substitution of Si4+ with 4H+.
Moreover this substitution increases the tetrahedron size,
but does not change its shape and orientation. As a result,
katoite-type oxygens are displaced, relative to grossular-
type oxygens, in a direction away from the position d along
the d-O vector. The existence of two O sites separated by
about 0.30 Å in garnet-hydrogarnet solid solutions and of an
associated high positional disorder were discussed in detail
by Armbruster & Lager (1989). These authors, on the basis
of a re-examination of the structure of a number of hydro-
garnets, evaluated the difference displacement parameters
2 U = U(O)-U(Si) along the tetrahedral d-O vector as a func-

** We dedicate this paper to the memory of the colleague and friend Luciano Ungaretti.

0935-1221/03/0015-0419 $ 3.60
DOI: 10.1127/0935-1221/2003/0015-0419 ˇ 2003 E. Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart
420 O. Ferro et al.
tion of OH– content, observing the highest value for hibschi-
te (“plazolite”) studied by Basso et al. (1983), where the two
O sites are about equally populated. Moreover, Armbruster
& Lager (1989) confirmed that the positional disorder in hy-
drogarnets can be described in terms of a “split-atom mod-
el” in which the two oxygen positions correspond to the sites
present in anhydrous grossular and Si-free katoite.
In this study we describe the occurrence and crystal struc-
ture of a new member of the hydrogrossular group, charac-
terized by a different 4H+/Si4+ ratio with respect to those re-
ported up to now in literature. The structural features of this
mineral are discussed in the context of the general structural
characteristics of the grossular-hydrogrossular solid solu-
tion.
Mineralogical description of katoite
Occurrence and paragenesis
The katoite studied was found in a quarry at Csódi Hill, near
Dunabogdány, a village in the Visegrád Mts., 50 km north of
Budapest, Hungary. Csódi Hill is a laccolite of dacite,
formed during the early volcanism of the Visegrád Mts, in
the Middle Miocene, 14.8 million years ago (Harangi,
1999). Fresh rock is bluish gray, while hydrothermal alter-
ation yielded a yellowish brown colour. The fine-grained
matrix embeds plagioclases, biotite and amphibole; garnets
are also scattered in the matrix. The xenoliths enclosed in
the dacite are characterized by the presence of brucite, lizar-
dite, clino-and orthocrysotile, deweylite, smectite, hydro-
grossular and calcite (Papp & Szakáll, 1999).
Katoite, rather rare, occurs in colourless crusts of small
octahedral crystals (side up to 0.11 mm) in the contact zone
(1-10 mm thick) between xenoliths and enclosing dacite.
Chemistry
Electron microprobe analysis was carried out in the wave-
length dispersive mode on an ARL-SEMQ instrument oper-
ated at 15 kV, 20 nA of probe current, using natural garnets
as standards. Mn, Ti, Zr, Na, K, P, V, Cr, S were analyzed,
but their content was lower than the detection limit. Due to
the limited quantity of the katoite sample, it was impossible
to perform TG analysis, hence the total H2O content was
calculated as the difference from 100% from the microprobe
analysis. Passaglia & Rinaldi (1984) determined the water
content of katoite from Montalto di Castro by both TG anal-
ysis and difference from 100% from the microprobe analy-
sis. They found a discrepancy between the two results of
less than 1%.
Table 1. Chemical composition and chemical formula (based on 12
oxygens) of katoite from Dunabogdány.
SiO2 16.54 Si 1.12
Al2O3 25.47 Al 2.03
CaO 40.76 Ca 2.95
FeO 0.55 Fe 0.03
H2O 16.68 H 7.51
Table 1 reports the chemical composition of katoite, ob-
tained by averaging seven point analyses, and the chemical
formula calculated on the basis of 12 oxygens. The formula
of the present katoite sample indicates a grossular content of
37%, which is a value very similar to one of the synthetic
members reported by Flint et al. (1941) and intermediate be-
tween that of “plazolite” by Basso et al. (1983) and katoite
by Passaglia & Rinaldi (1984).
X-ray data and physical properties
The X-ray powder pattern of katoite was obtained with a
Gandolfi camera. As discussed by Passaglia & Rinaldi
(1984), the increase of OH– content in hydrogrossulars cor-
responds to an increase of the dhkl and to the progressive ap-
pearance of the (211), (321), and (220) diffraction lines, the
latter being the distinctive character of the katoite members
with respect to the hibschite ones. All these lines are present
in our powder pattern. The refinement of the unit cell pa-
rameter based on 22 diffraction lines yielded a = 12.309(2)
Å.
Under the polarizing microscope, katoite crystals are col-
ourless. The refractive index, determined by the Becke line
method, is n = 1.651(1). This value is intermediate between
that of katoite (n = 1.632; Passaglia & Rinaldi, 1984) and
that of Crestmore hibschite (n = 1.675; Pabst, 1942) and
closely fits the linear correlation between refractive index
and Gr% reported in Fig. 1.
Structural description of katoite
Experimental
The quality of katoite crystals was tested by preliminary
precession photographs. No streaks, splitting or additional
reflections were observed and the spots were very neat and
sharp.
An octahedral katoite crystal (side of 0.11 mm) was
mounted on a Siemens 4-circle diffractometer P4R with ro-
tating anode X-ray generator (MoK [ radiation). Two data
collections were performed at room temperature and 200 K
respectively in the K scan mode (scan angle = 0.76°). Unit
Table 2. Crystal and refinement data for katoite at 200 K and room
temperature in the unsplit- and split-O models.
200 K Room T
a (Å) 12.270(1) 12.286(1)
Measured reflections 2425 7705
Unique reflections 331 1002
R(int) % 3.35 5.25
Secondary extinction coefficient 0.0008(2) 0.0015(5)
(unsplit-O model)
Secondary extinction coefficient 0.0007(2) 0.0011(5)
(split-O model)
Reflections with Fo > 4 c (Fo) 299 744
R(Fo), Rw (Fo2)% (unsplit-O model) 1.97, 4.45 4.01, 11.49
R(Fo), Rw (Fo2)% (split-O model) 2.34, 4.68 4.56, 12.42
 New occurrence of katoite 421

Table 3. Atomic coordinates, site occupancy factors and thermal parameters for katoite at 200 K in the unsplit- and split-O models.

Atom x y z Occ. U11 U22 U33 U23 U13 U12 Ueq/Uiso

(x104)
Unsplit-O model
Ca 1/8 0 1/4 0.972(4) 108(2) 87(1) 87(1) 23(1) 0 0 94(1)
Al 0 0 0 1.002(6) 137(2) 137(2) 137(2) -36(1) -36(1) -36(1) 137(1)
Si 3/8 0 1/4 0.364(4) 59(5) 65(3) 65(3) 0 0 0 63(2)
O .03237(8) .04982(7) .64450(7) 1 298(4) 173(4) 187(4) -73(3) 120(3) -72(3) 219(2)
H .075(4) .061(4) .655(3) 0.636 900

Split-O model
Ca 1/8 0 1/4 0.972(4) 108(2) 87(2) 87(2) 23(1) 0 0 94(1)
Al 0 0 0 1.006(6) 136(3) 136(3) 136(3) -37(2) -37(2) -37(2) 136(3)
Si 3/8 0 1/4 0.368(4) 63(6) 70(4) 70(4) 0 0 0 68(3)
O1 .0439(3) .0436(2) .6528(2) 0.37 84(5)
O2 .0250(2) .0537(2) .6393(2) 0.63 152(4)
H .081(3) .060(3) .656(3) 0.63 380(118)




 SHELX-97 computer program (Sheldrick, 1997). The

 oms were used for the corresponding sites. The refinements
 were done in the space group Ia-3d. The secondary extinc-
 
 tional disorder of oxygen atoms, both unsplit- and split-O
 .'

 finement of its atomic coordinates and isotropic thermal pa-

        
*U
Fig. 1. Gr% vs. refractive index for natural members of the
Ca3Al2(SiO4)3–Ca3Al2(O4H4)3 series.
KD: katoite from Dunabogdány; 1: katoite (Passaglia & Rinaldi,
1984); 2: plazolite from Crestmore (Pabst, 1942); 3: hibschite from
Bugriver (Nalivkina, 1960); 4: hibschite from Nikortzminda (Bely-
ankin & Petrov, 1941); 5: hydrogrossular from Hatrurim formation
(Gross, 1977); 6: hydrogrossular from Dunn Mt. (Hutton, 1943); 7:
South African “jade”, Bushveld complex (Frankel, 1959).
cell dimensions were calculated from least-square refine-
ments of the d values obtained for 57 diffractions (Table 2).
Intensity data were collected up to ’ = 55° in the room
temperature experiment and to 34.5° in the low temperature
experiment due to the constructional limits of the cryostat.
X-ray intensity data was collected in one octant of the Ewald
sphere and corrected for the Lorentz polarisation effect and
for absorption (the latter correction was applied only to the
room temperature data) by the method of North et al.
(1968). The details of the two data collections are reported
in Table 2.
TEM investigations were carried out on a powdered sam-
ple with a JEOL 2011 operating at 200 KeV. The obtained SA-
ED, collected on different crystals, showed well defined and
sharp spots indicating no structural disordering at this scale.
Structural refinements
Least-square refinements were carried out using the
atomic scattering curves for neutral Ca, Al, Si, O and H at-
tion coefficient was refined (Table 2). To evaluate the posi-
models were used in the refinements at both temperatures.
The hydrogen position was detected in the difference Fo-
urier map of only the low-temperature refinements. The re-
rameter yielded the values reported in Table 3 for unsplit-
and split-O models, respectively.
Anisotropic temperature factors for all atoms, except hy-
drogen, were used in the unsplit-O model refinements,
whereas isotropic temperature factors were used for oxygen
and hydrogen atoms in the split-O model.
The two oxygen positions in the split-O model were de-
termined by the difference Fourier map and were consistent
with those reported by Armbruster & Lager (1989) for kato-
ite, studied by Sacerdoti & Passaglia (1985), and “plazoli-
te”, studied by Basso et al. (1983). They are labelled O1 and
O2 respectively, where O1 is bonded to Si and O2 to H. The
final R and RW agreement factors of the four refinements are
reported in Table 2. Atomic coordinates, s.o.f. and thermal
parameters for both models are reported in Table 3, inter-
atomic distances and angles for the unsplit- and split-O
models are reported in Tables 4 and 5.
Results and discussion
The unit cell parameters of katoite resulted 12.270(1) and
12.286(1) Å for the experiment at 200 K and room T respec-
tively (Table 2); the difference between the two values (0.015)
is consistent with that found by Lager et al. (1987) for Si-free
hydrogrossular in the same temperature range (0.016).
Passaglia & Rinaldi (1984) reported a linear relationship
between the Gr% versus unit cell parameter for a series of nat-
422 O. Ferro et al.

Table 4. Selected interatomic distances (Å) and angles (°) in the unsplit-O model for members of the Ca3Al2(SiO4)3–Ca3Al2(O4H4)3 series.

Grossular 1 “Plazolite” 2 This work Katoite 3 Si-free

200 K katoite 4
Tetrahedron
4 d-O 1.645(1) 1.775(1) 1.828(1) 1.892(2) 1.962(1)
2 O(1)-O(2) 2.567(2) 2.775(2) 2.863(2) 2.969(2) 3.076(1)
4 O(1)-O(3) 2.745(1) 2.957(2) 3.044(2) 3.147(2) 3.267(1)
2 O(1)-d-O(2) 102.53(8) 102.88(7) 103.10(5) 103.4(1) 103.19(4)
4 O(1)-d-O(3) 113.05(4) 112.86(4) 112.75 (3) 112.6(1) 112.70(3)
Octahedron
6 Al-O 1.924(1) 1.927(1) 1.9171(9) 1.902(2) 1.906(1)
6 O(1)-O(4) 2.756(2) 2.701(2) 2.660(2) 2.607(2) 2.578(1)
6 O(1)-O(5) 2.686(2) 2.750(2) 2.761(1) 2.769(3) 2.810(1)
6 O(1)-Al-O(4) 91.46(5) 88.98(4) 87.87(4) 86.6(1) 85.80(3)
6 O(1)-Al-O(5) 88.54(5) 91.02(4) 92.13(4) 93.4(1) 94.92(3)
Dodecahedron
4 Ca(1)-O(4) 2.319(1) 2.383(1) 2.4037(9) 2.429(1) 2.462(1)
4 Ca(2)-O(4) 2.490(1) 2.503(1) 2.4996(9) 2.497(3) 2.511(1)
2 O(1)-O(2) 2.567(2) 2.775(2) 2.863(2) 2.969(2) 3.076(1)
4 O(1)-O(4) 2.756(2) 2.701(2) 2.660(2) 2.607(2) 2.578(1)
4 O(1)-O(7) 3.450(1) 3.573(1) 3.6161(6) 3.669(3) 3.738(1)
4 O(4)-O(6) 2.973(2) 3.011(2) 3.021(2) 3.037(3) 3.059(1)
2 O(4)-O(7) 2.866(2) 2.931(2) 2.951(2) 2.977(3) 3.031(1)
2 O(7)-O(8) 4.121(2) 4.092(2) 4.063(2) 4.030(3) 4.023(1)
2 O(1)-Ca(2)-O(2) 67.20(6) 71.23(5) 73.10(5) 75.4(1) 77.33(1)
4 O(1)-Ca(2)-O(4) 69.84(4) 67.08(5) 65.68(4) 63.9(1) 62.45(2)
4 O(1)-Ca(2)-O(7) 91.61(4) 93.95(3) 95.01(3) 96.3(1) 97.46(2)
4 O(4)-Ca(2)-O(6) 76.28(6) 76.03(4) 76.04(3) 76.1(1) 75.91(2)
2 O(4)-Ca(2)-O(7) 70.27(6) 71.68(5) 72.36(4) 73.2(1) 74.27(2)
2 O(7)-Ca(2)-O(8) 111.66(6) 109.68(5) 108.75(4) 107.6(1) 106.36(2)
H-distances
O-H 0.928(1) 0.56(4) 0.68 0.65(1)
d-H 1.713(1) 1.52(4) 1.24 1.43(1)
Al-H 2.315(1) 2.24(4) 2.30 2.29(1)
Ca(1)-H 3.239(1) 2.82(3) 2.62 2.86(1)
Ca(2)-H 2.398(1) 2.52(3) 2.96 2.75(1)

1Novak & Gibbs (1971); 2Basso et al. (1983); 3Sacerdoti & Passaglia (1985); 4Lager et al. (1987a) (300 K,

X-ray with hydrogen refinement).

ural and synthetic hydrogarnets. The a value obtained by the
single-crystal experiment at room T is consistent with 39%
of Gr component, while that obtained by powder diffraction
is consistent with a Gr% = 36. Both Gr% values agree well
with that obtained by the chemical analysis (Gr% = 37).
The results of the structure refinements at room tempera-
ture will not be discussed in detail below, since they do not
significantly differ from those at 200 K. The low-T data will
be discussed on the basis of the two different models (un-
split- and split-O) used for the structure refinements.
Unsplit-O model
In this refinement the Si site occupancy resulted in a value of
36%, in close agreement with the chemical data; as a conse-
quence, the H site occupancy was fixed to 64%.
As discussed by Armbruster & Lager (1989), the substi-
tution of 4H+ with Si4+ in hydrogarnets induces an increase
in the mean-square displacement amplitude of oxygen in
the direction of the d-O vector. This behaviour is con-
firmed by the oxygen ADP factors determined in this work
(Table 3 and Fig. 2) and by the value of 2 U = U (O) – U
(Si) (difference displacement parameter evaluated in the d-
O direction), which is 335· 10-4 Å2. This value is quite simi-
lar to that obtained by Armbruster & Lager (1989) for “pla-
zolite” (338· 10-4 Å2), but it better fits the theoretical value
predicted by the same authors on the basis of an equation
(Chandrasekhar & Bürgi, 1984) which correlates the 2 U =
U(O) – U(Si) with the d-O distance along the hydrogrossu-
lar series.
Table 4, which reports the structural features of some
terms of the Ca3Al2(SiO4)3–Ca3Al2(O4H4)3 series, ordered
by decreasing Si content, allows a comparative description
of how the different polyhedra of the garnet structure modi-
fy along the series. Codes proposed by Novak & Gibbs
(1971) have been used throughout.
 New occurrence of katoite 423

Table 5. Selected interatomic distances (Å) of katoite from Dunabogdány refined in the split-O model compared with other members of the
Ca3Al2(SiO4)3–Ca3Al2(O4H4)3 series.
Grossular1 “Plazolite” 2 This work 200 K Katoite2 Si-free katoite3
a = 11.845(1) Å a = 12.174(3) Å a = 12.286(1) Å a = 12.358(3) Å a = 12.5695(1) Å
Tetrahedron
4 d-O1 1.645(1) 1.621 1.643(3) 1.695
4 d-O2 1.936 1.946(2) 1.918 1.962(1)
2 O1(1)-O1(2) 2.567(2) 2.593 2.615(6) 2.717
2 O2(1)-O2(2) 3.000 3.019(4) 3.000 3.076(1)
4 O1(1)-O1(3) 2.745(1) 2.675 2.716(6) 2.795
4 O2(1)-O2(3) 3.197 3.253(3) 3.197 3.267(1)
2 O1(1)-d-O1(2) 102.53(8) 106.18 105.5(2) 106.46
2 O2(1)-d-O2(2) 102.86 101.8(1) 102.87 103.19(4)
4 O1(1)-d-O1(3) 113.05(4) 111.14 111.5(1) 111.00
4 O2(1)-d-O2(3) 112.87 113.6(1) 112.87 112.70(3)
Vtetrahedron (O1) (Å3) 2.26 2.27
Vtetrahedron (O2) 3.73 3.84
Octahedron
6 Al-O1 1.924(1) 2.014 2.023(3) 1.933
6 Al-O2 1.843 1.857(2) 1.911 1.906(1)
6 O1(1)-O1(4) 2.756(2) 2.957 2.716(6) 2.880
6 O2(1)-O2(4) 2.579 2.470(4) 2.579 2.578(1)
6 O1(1)-O1(5) 2.686(2) 2.735 2.755(5) 2.580
6 O2(1)-O2(5) 2.821 2.775(3) 2.821 2.810(1)
6 O1(1)-Al-O1(4) 91.46(5) 94.46 94.1(1) 96.29
6 O2(1)-Al-O2(4) 83.3(1) 84.88 85.80(3)
6 O1(1)-Al-O1(5) 88.54(5) 85.53 85.9(1) 83.71
6 O2(1)-Al-O2(5) 96.7(1) 95.12 94.92(3)
Voctahedron (O1) (Å3) 9.49 10.94
Voctahedron (O2) 8.38 9.15
Dodecahedron
4 Ca(1)-O1(4) 2.319(1) 2.442 2.451(3) 2.484
4 Ca(1)-O2(4) 2.335 2.380(2) 2.419 2.462(1)
4 Ca(2)-O1 (4) 2.490(1) 2.589 2.611(3) 2.726
4 Ca(2)-O2(4) 2.421 2.434(2) 2.456 2.511(1)
2 O1(1)-O1(2) 2.567(2) 2.593 2.615(6) 2.717
2 O2(1)-O2(2) 3.000 3.061(4) 3.000 3.076(1)
4 O1(1)-O1(4) 2.756(2) 2.957 2.716(6) 2.880
4 O2(1)-O2(4) 2.579 2.469(4) 2.579 2.578(1)
4 O1(1)-O1(7) 3.450(1) 3.557 3.583(5) 3.669
4 O2(1)-O2(7) 3.664 3.645(4) 3.660 3.738(1)
4 O1(4)-O1(6) 2.973(2) 3.141 3.160(6) 3.404
4 O2(4)-O2(6) 2.965 2.933(4) 2.965 3.059(1)
2 O1(4)-O1(7) 2.866(2) 2.854 2.900(6) 3.056
2 O2(4)-O2(7) 2.958 2.994(4) 2.958 3.031(1)
2 O1(7)-O1(8) 4.121(2) 4.375 4.397(6) 4.547
2 O2(7)-O2(8) 3.945 3.856(4) 3.945 4.023(1)
2 O1(1)-Ca(2)-O1(2) 67.20(6) 64.14 64.48(9) 66.30
2 O2(1)-Ca(2)-O2(2) 76.60 78.7(1) 76.60 77.33(1)
4 O1(1)-Ca(2)-O1(4) 69.84(4) 71.92 71.53(8) 66.92
4 O2(1)-Ca(2)-O2(4) 63.86 61.71(9) 63.86 62.45(2)
4 O1(1)-Ca(2)-O1(7) 91.61(4) 89.92 90.05(5) 89.38
4 O2(1)-Ca(2)-O2(7) 97.39 98.41(6) 97.39 97.46(2)
4 O1(4)-Ca(2)-O1(6) 76.28(6) 77.18 77.19(6) 81.44
4 O2(4)-Ca(2)-O2(6) 74.89 75.06(7) 74.89 75.91(2)
2 O1(4)-Ca(2)-O1(7) 70.27(6) 66.89 67.46(8) 68.16
2 O2(4)-Ca(2)-O2(7) 74.03 75.89(9) 74.03 74.27(2)
2 O1(7)-Ca(2)-O1(8) 111.66(6) 115.28 114.68(8) 112.97
2 O2(7)-Ca(2)-O2(8) 106.82 104.73(9) 106.82 106.36(2)
Vdodecahedron (O1) (Å3) 23.89 27.57
Vdodecahedron (O2) 23.34 25.76
H-distances in the “O2” model
O2-H 0.92(1) 0.72(4) 0.701(8) 0.65(1)
d-H 1.21 1.47(4) 1.41 1.43(1)
Al-H 2.26 2.28(4) 2.23 2.29(1)
Ca(1)-H 2.85 2.87(4) 2.25 2.86(1)
Ca(2)-H 2.98 2.57(4) 3.08 2.75(1)
1Novak & Gibbs (1971); 2Basso et al. (1983); 2Armbruster & Lager (1989); 3Lager et al. (1987a) (300 K, X-ray with hydrogen refinement).
424 O. Ferro et al.



 

.'







       
Fig. 2. Tetrahedron geometry for katoite from Dunabogdány. The
mean-square displacement amplitude of oxygen atom as obtained by
the unsplit-O model refinement.
In general, linear correlations may be observed between
the degree of hydration and the polyhedra geometric param-
eters, from which our sample does not deviate. In particular,
the d-O and the O-O tetrahedron distances increase, while
the Y-O octahedron distance remains almost constant as a
function of the OH content. The octahedron deformation is
achieved through the shortening of the O(1)-O(4) edge, the
lengthening of O(1)-O(5), and the corresponding variations
in the relative bond angles (see Table 4). The triangular do-
decahedron shows the same deformations described by Sa-
cerdoti & Passaglia (1985): the shortest X-O distance no-
ticeably increases; the longest one is almost constant; the
O(1)-O(2) edge, shared with the tetrahedron, shows the
strongest increase along the series; the O(1)-O(4) edge,
shared with the octahedron, decreases; the unshared edges
and those shared with the other dodecahedra show minor
variations as a function of the OH content.
As reported by Basso et al. (1983) and Sacerdoti & Pas-
saglia (1985), another structural feature that undergoes a
strong increase with the degree of hydration, is the polar co-
ordinate of the oxygen atom ” ox = arctan (yox/xox). The ” ox
value for our sample, 56.99, is intermediate between those
of “plazolite” (54.83) and katoite (59.43). On the basis of
the linear correlation reported in Fig. 3, this value corre-
sponds to 7.3 OH– in the chemical formula and to a Gr% of
39, closely agreeing with the results of the chemical analysis
(7.51 OH–) and the above reported value derived from the
unit cell parameter.
Split-O model
Column III of Table 5 gives selected interatomic bond dis-
tances derived from the refinement of our katoite sample in
2+
Fig. 3. Polar coordinates of the oxygen atom obtained by the unsplit-
O model refinement vs. the OH content.
KD: katoite from Dunabogdány; 1: grossular (Novak & Gibbs,
1971); 2: plazolite (Basso et al., 1983); 3: katoite (Sacerdoti & Pas-
saglia, 1985); 4: Si-free katoite (Lager et al., 1987b).
the split-O model. The resulting Si occupancy in this refine-
ment model was 37% (Table 3), in excellent agreement with
the value obtained from the unsplit refinement and from the
other physical and chemical data. Due to the short distance
between the two oxygen sites O1 and O2 (0.31 Å) and to the
consequent high correlations between their occupancy fac-
tors, the site occupancy of the O1 oxygen was fixed to the Si
occupancy value, whereas that of O2 and H were fixed to the
difference from 100% (63%).
Table 5 also reports the polyhedra geometry of the two
end-member grossular (Novak & Gibbs, 1971) and Si-free
katoite (Lager et al., 1987a) and the two structural configu-
rations obtained from the split-O model refinements of the
intermediate members of the hydrogrossular series (“plazo-
lite” and katoite by Armbruster & Lager (1989) and katoite
of this work). It is worth noting that the comparison between
the geometrical features of these structures must take into
account the differences in the cell parameters, also reported
in Table 5.
At a first glance of Table 5, tetrahedron and octahedron
geometrical parameters of the two “O1” and “O2” configura-
tions could be consistent with a description of the structure
of the intermediate members based on the coexistence of
two structural models: the former similar to grossular and
the latter similar to Si-free katoite. For instance, the Si for H
substitution brings about the increase of the volume of the
tetrahedron and reduces that of the octahedron. However,
the analysis of the X-ray diffraction peaks and the results of
the TEM study rule out the possibility to describe the katoite
structure with the presence of domains of the two end-mem-
ber configurations (grossular and Si-free katoite). A further
evidence of the homogeneity of katoite structure comes
from the following crystallographic argument: if domains of
the two end-members were present, the description of the
tetrahedron geometrical features – based on the split-O
model coordinates and the intermediate katoite unit cell –
 New occurrence of katoite
would provide erroneous and unexpected bond distances,
contrary to that observed in Table 5.
In conclusion, katoite structure can be described with a
random distribution in the unit cell of two types of tetrahe-
dra (SiO4 and O4H4) whose volume differences are absorbed
by the less rigid octahedra and dodecahedra present in the
garnet structure, which share oxygens with the two different
tetrahedra. Hence, the structural features of the octahedron
and dodecahedron reported in Table 5 correspond to a de-
scription based on two extreme configurations which proba-
bly are not present in the real structure of katoite.
Hydrogen position
Fig. 2 displays the atomic environment around d position of
the Dunabogdány katoite, as obtained by the unsplit-O re-
finement. The O-H vectors are located outside the tetrahe-
dral faces, as observed for “plazolite” (Basso et al., 1983)
and Si-free katoite (Lager et al., 1987a).
The positional parameters of the H atom in the split- and
unsplit-O refinements are very similar to each other (Table
3) and closer to the ones of Si-free katoite (X-ray refine-
ment) rather than to the other hydrogrossulars reported by
Armbruster & Lager (1989). The O-H bond distances are
reported in Table 4 and 5 for the unsplit- and split-O mod-
els, respectively. As in the unsplit- X-ray structure refine-
ments of other hydrogrossulars, the O-H distance in this
katoite is very short (0.56(4)); however it significantly in-
creases in the split-O model (0.72(4)). Moreover, all the
distances between hydrogen and the other cations deter-
mined in the “O2” configuration, are very similar to those
found by Lager et al. (1987a) for Si-free katoite (Table 5),
further confirming the strong similarity between the O4H4
tetrahedra of the katoite from Dunabogdány and of the Si-
free end-member.
Acknowledgement: Simona Bigi is acknowledged for the
assistance during HRTEM experiments and Alberto Alberti
for the critical reading of the manuscript. The paper was im-
proved by the constructive reviewing by F. Mazzi and Th.
Armbruster. The katoite-bearing sample was collected by
Csaba Ott (Budapest). The “Consiglio Nazionale delle Ri-
cerche” of Italy is acknowledged for financing the electron
microprobe laboratory at the Dipartimento di Scienze della
Terra of the University of Modena and Reggio Emilia. The
single-crystal X-ray diffraction experiments were carried
out at Centro Interdipartimentale Grandi Strumenti of the
University of Modena and Reggio Emilia. Italian MIUR and
CNR are acknowledged for financial support.
References
Armbruster, Th. (1995): Structure refinement of hydrous andradite,
Ca3Fe1.54Mn0.20 Al0.26(SiO4)1.65(O4H4)1.35, from the Wessel mine,
Kalahari manganese field, South Africa. Eur. J. Mineral., 7,
1221-1225.
Armbruster, Th. & Lager, G.A. (1989): Oxygen disorder and the hy-
drogen position in garnet-hydrogarnet solid solutions. Eur. J.
Mineral., 1, 363-369.
425
Basso, R., Della Giusta, A., Zefiro, L. (1981): A crystal chemical
study of a Ti-containing hydrogarnet. N. Jb. Mineral. Mh., 1981,
230-236.
––,––,–– (1983): Crystal structure refinement of plazolite: a highly
hydrated natural hydrogrossular. N. Jb. Mineral. Mh., 1983, 251-
258.
Basso, R., Cimmino, F., Messiga, B. (1984a): Crystalchemistry of
hydrogarnets from three different microstructural sites of a basal-
tic metarodingite from Voltri massif (Wester Liguria, Italy). N.
Jb. Mineral. Abh., 148, 246-258.
––,––,–– (1984b): Crystal chemical and petrological study of hydro-
garnets from a Fe-gabbro metarodingite (Gruppo di Voltri, West-
er Liguria, Italy). N. Jb. Mineral. Abh., 150, 247-258.
Belyankin, D.S. & Petrov, V.P. (1941): The grossularoid group (hib-
schite, plazolite). Amer. Mineral., 26, 450-453.
Chandrasekhar, K. & Bürgi, H.B. (1984): Dynamic processes in
crystals examined through difference vibrational parameters 2 U:
the low-spin-high-spin transition in tris(dithiocarbamato)iron(I-
II) complexes. Acta Cryst., B40, 387-397.
Cohen-Addad, C., Ducros, P., Bertaut, E.F. (1967): Etude de la susti-
tution du groupement SiO4 par (OH)4 dans les composés
Al2Ca3(OH)12 et Al2Ca3(SiO4)2.16(OH)3.36 de type grenat. Acta
Cryst., 23, 220-230.
Flint, E.P., McMurdie, H.F., Wells, L.S. (1941): Hydrothermal and
X-ray studies of the garnet-hydrogarnet series and the relation-
ship of the series to hydration products of Portland cement. J.
Res. Natl. Bur. Stds., 26, 13-33.
Frankel, J.J. (1959): Uvarovite garnet and South African jade (hy-
drogrossular) from the Bushveld complex. Transvaal. Amer. Min-
eral., 44, 565-591.
Gross, S. (1977): The mineralogy of the Hatrurim formation, Israel.
Geol. Surv. Israel, Bull. No. 70, 1-80.
Harangi, Sz. (1999): Geochemistry and petrogenesis of the volcanic
rocks from Csódi Hill (Dunabogdány, Visegrád Mts., Hungary).
Topogr. Mineral. Hung. (Miskolc), 6, 59-85. (Hungarian with
English abstract)
Hsu, L.C. (1980): Hydration and phase relations of grossular-spes-
sartine garnets at PH2O= 2Kb. Contrib. Mineral. Petrol., 71, 407-
415.
Hutton, C.O. (1943): Hydrogrossular, a new mineral of the garnet-
hydrogarnet series. Trans. Roy. Soc. New Zealand, 73, 174-180.
Lager, G.A., Armbruster, Th., Faber, J. (1987a): Neutron and X-ray
diffraction study of hydrogarnet Ca3Al2(O4H4)3. Amer. Mineral.,
72, 756-765.
Lager, G.A., Rossman, G.R., Rotella, F.J., Schultz, A.J. (1987b):
Neutron-diffraction structure of a low-water grossular at 20K.
Amer. Mineral, 72, 766-768.
Lager, G.A., Armbruster, Th., Rotella, F.J., Rossman, G.R. (1989): OH
substitution in garnets: X-ray and neutron diffraction, infrared, and
geometric-modeling studies. Amer. Mineral., 74, 840-851.
Nalivkina, É. B. (1960): Hibschite from near the Bug River. Zapiski
Vses. Mineralog. Obshch. (Mem. All-Union Min. Soc.), 89, 714-
718 (in Russian) (Mineral. Abs., 16, 454).
North, A.C.T., Phillips, D.C., Mathews F.S. (1968): A semi-empiri-
cal method of absorption correction. Acta Cryst., A 24, 351-359.
Novak, G.A. & Gibbs, G.V. (1971): The crystal chemistry of silicate
garnets. Amer. Mineral., 56, 791-825.
Pabst, A. (1942): Reexamination of hibschite. Amer. Mineral., 27,
783-792.
Papp, G. & Szakáll, S. (1999): Minerals of serpentine-bearing xeno-
liths from Csódi Hill (Dunabogdány, Visegrád Mts., Hungary).
Topogr. Mineral. Hung. (Miskolc), 6, 103-125. (Hungarian with
English abstract)
426 O. Ferro et al.

Passaglia, E. & Rinaldi, R. (1984): Katoite, a new member of the Sheldrick, G.M. (1997): SHELX-97: Program for the solution and
Ca3Al2(SiO4)3-Ca3Al2(OH)12 series and a new nomenclature for refinement of crystal structures. Siemens Energy and Automa-
the hydrogrossular group of minerals. Bull. Minéral., 107, 605- tion, Madison, WI.
618.
Sacerdoti, M. & Passaglia, E. (1985): The crystal structure of katoite Received 9 November 2001
and implications within the hydrogrossular group of minerals. Modified version received 29 May 2002
Bull. Minéral., 108, 1-8. Accepted 25 June 2002