High-pressure and high-temperature polymorphism of Na₂CO₃ up to 10 GPa

We report two new high-pressure polymorphs of sodium carbonate, ɛ and ɛ-II, stable over ∼2 GPa and 11 GPa, respectively. The ɛ polymorph exhibits a doubling of the c unit-cell parameter relative to the γ and β phases, crystallizes in space group Cc, and has a bulk modulus of 47.6 (8) GPa, comparable to that of the γ phase. Its stability is confirmed using in situ single-crystal and large-volume press (LVP) experiments. Phase transitions observed over a broad pressure–temperature range allow us to refine the topology of the Na₂CO₃ phase diagram. By integrating precise single-crystal constraints with LVP observations, we propose an updated phase diagram that shows an expanded stability field for the incommensurate γ phase, and reveals a wide stability field for the ɛ phase at conditions relevant for the Earth's mantle. These results provide improved experimental constraints for the high-pressure behavior of sodium carbonate.

1. Introduction

Sodium carbonate (Na₂CO₃) draws crystallographic attention due to its complex polymorphism and modulated structure. The historical significance of this phase is due to the incommensurate structure of the room-temperature polymorph γ, whose structural solution has been complicated to finalize. Brouns & Visser (1964) reported the first description of the γ lattice underlining the necessity to use four indexes to describe the satellite reflections observed in diffraction patterns. It has been the first structure to be refined as a modulated phase using a harmonic approximation (De Wolff, 1974; Van Aalst et al., 1976). Further work by Dusek et al. (2003) highlighted the importance of incorporating multiple modulation waves in the structure solution, distinguishing between harmonic and anharmonic descriptions.

The interest in Na₂CO₃ arises also from the wide number of temperature-driven phase transitions observed at atmospheric pressure, where four different polymorphs are stable (Harris & Salje, 1992):

δ (P2/a) → 130 K → γ (C2/m average) → 620 K → β (C2/m) → 763 K → α (P63/mmc).

The δ phase structure was initially described by de Pater & Helmholdt (1979) and later confirmed by Harris & Salje (1992). Its structure was then reported by Dusek et al. (2003), in comparison with the γ structure. While γ and δ are structurally similar, δ is a commensurate phase with lower modulation amplitudes and reduced atomic displacement parameters. The γ–β phase transition leads to a monoclinic structure, while at higher temperatures, the β–α phase transition represents a ferro­elastic phase transition (Harris & Dove, 1995; Swainson et al., 1995; Harris et al., 1997). More recently, Ballirano (2011) provided an in-depth structural characterization of Na₂CO₃ phases and their temperature-driven transitions, using in situ experimental techniques. The high-temperature behavior of Na₂CO₃ has been also presented as a model system for lattice melting (Swainson et al., 1995).

De Wolff & Tuinstra (1986) provided detailed structural descriptions of the α, β and γ phases, emphasizing the role of NaO₆ octahedral columns and their evolution under stress and modulation. They described the Na₂CO₃ structure as composed of sodium face-sharing octahedra columns (hosting Na1 and Na2 atoms) parallel to the hexagonal axis of the structure, interconnected by the CO₃ groups. The space between the CO₃ groups and the columns hosts Na3 atoms, placed in between the columns and the CO₃ groups. Dusek et al. (2003) underlined the importance of the Na3 site, whose coordination environment and bond dynamics appear to drive phase transitions. In the α phase, long Na–O distances are stabilized by high thermal motion; in the β phase, shear stress brings atoms closer, with a vertical displacement of the Na columns and a reduction in symmetry; and in the γ phase, strong modulation results in variable bonding configurations. The evolving bonding behavior of this Na site possibly represents a key mechanism behind the polymorphic transitions.

From the geological perspective, the relevance of Na₂CO₃ becomes more pronounced at HP-HT conditions. Early studies (e.g. Shatskiy et al., 2013) proposed a phase diagram based on energy-dispersive techniques. More robust findings from multianvil (MA) and diamond anvil cell (DAC) experiments (Gavryushkin et al., 2019) based on the powder diffraction technique, suggested the stability of γ up to 12 GPa in MA experiments, with a phase transition occurring near 26 GPa. Interestingly, the γ structure remains stable up to 25 GPa in DAC experiments, suggesting discrepancies due to resolution limits in MA data, which often fails to distinguish between β and γ phases. At pressures above 12 GPa and high temperature (Gavriuskin et al., 2019), the γ structure is replaced by a hexagonal phase, which is demonstrated to be the thermodynamically stable phase from those pressures to 25 GPa, where it is replaced by a higher-pressure monoclinic polymorph (Gavriuskin et al., 2019).

Given the complexity of the low-pressure polymorphism, the crystallographic interest for the presence of incommensurate structure, and also the geological importance of this system, a comprehensive understanding of its phase diagram, especially in the pressure interval 0–10 GPa, is essential for both crystallographic modeling and geophysical applications.

We report an in situ high-pressure and high-temperature study to assess the thermo-elastic properties of Na₂CO₃ and its stability phase diagram. We also report two new high-pressure structures; one of these appears to be an additional stable polymorph.

2. Methods

Single crystals of sodium carbonate were crystallized from melting pure reagent sodium carbonate powder in a platinum crucible, reaching a temperature of 1124 K. Careful crystal selection under the microscope allowed the identification of single domains that were then used for different in situ single-crystal X-ray diffraction (SC-XRD) measurements (Table 1 for details).

Table 1
Experimental details for different runs

Run	Setup	Pressure medium	P (GPa)	T (K)	Beamline
1	DAC	Helium	0–38.37 (5)	RT	ID15b, ESRF
2	DAC	Silicone oil	0–8.55 (5)	RT	ID09, ESRF
3	Resistively heated DAC	Silicone oil	Fig. 1(a)		ID15b, ESRF
4	Resistively heated DAC	Silicone oil	Fig. 1(a)		ID15b, ESRF
5	LVP	–	Fig. 1(b)		ID06-LVP, ESRF
6	LVP	–	Fig. 1(b)		ID06-LVP, ESRF

One experiment (Run 1) was performed in a membrane Boehler–Almax design DAC with 300 µm culet diamonds and stainless-steel gasket. The cell has been loaded with a single crystal of Na₂CO₃ and a ruby sphere, with helium as pressure transmitting medium. SC-XRD data were collected at ID15b beamline at the European Synchrotron Radiation facility (ESRF, Grenoble, France). The beamline was equipped with a Mar555 image plate detector and data were collected with a wavelength of 0.41105 Å (Merlini & Hanfland, 2013). Pressure was increased up to 38.37 (5) GPa in small steps and for each step pressure was calibrated using ruby fluorescence (Mao et al., 1986).

A second experiment (Run 2) was performed using a 600 µm culet DAC with stainless-steel gasket. In this case, a single crystal of Na₂CO₃ and the ruby were loaded in the cell along with silicone oil as pressure medium. Data were collected up to 8.55 (5) GPa at the ID09 beamline (ESRF, Grenoble, France). For beamline details see Merlini & Hanfland (2013). A Mar555 detector and a wavelength of ∼0.41 Å were used.

High-pressure and high-temperature experiments were performed using an internal resistively heated Almax DAC equipped with 600 µm culet diamonds, Inconel gaskets and silicone oil as pressure medium. One experiment (Run 3) involved two high-temperature ramps, one around 4 GPa and a second one around 7 GPa. The maximum pressure reached was 9.00 (5) GPa, while the high-temperature ramps reached 690 K and 665 K, respectively [Fig. 1(a)]. Pressure was measured using samarium-doped SrB₄O₇ fluorescence while temperature was measured using a thermocouple placed in direct contact with one diamond. Data were collected at ID15b (ESRF, Grenoble, France) with a Mar555 detector and a wavelength of 0.41105 Å.

Figure 1 Pressure–temperature plots for (a) Run 3 and Run 4 and (b) Run 5 and Run 6.

Another HP-HT experiment (Run 4) was conducted at the same beamline with an Eiger2X 9M CdTe detector [further details on the beamline setup given by Poręba et al. (2022)]. In this case, the maximum pressure reached was 2.44 (5) GPa and the maximum temperature 1020 K. Pressure was derived from the pressure–volume–temperature (PVT) equation of state (Angel et al., 2017) of a quartz crystal placed in the cell along with the sample, while temperature was measured with a thermocouple. The pressure–temperature paths followed during the two experiments are given in Fig. 1(a), while information on the P/T conditions of the phase transitions is available as supporting information, S1.

SC-XRD data were indexed and integrated with CrysAlis PRO RED software (Rigaku Oxford Diffraction, 2019). For different pressure/temperature points, structural solution and refinement were done with Jana2006 software (Petříček et al. 2014), using the SuperFlip charge-flipping algorithm (Palatinus & Chapuis 2007).

The last experiments (Run 5 and Run 6) were performed with the large-volume press (LVP) available at the ID06-LVP beamline (ESRF, Grenoble, France).

Both experiments used the standard 10/5 assembly. Previously dehydrated sodium carbonate crystals were finely ground along with platinum powder in an agate mortar and then pressed into a pellet of 2.4 mm diameter and 1.8 mm thickness. Sample and calibrant (platinum in Run 5, gold in Run 6) were placed directly into the graphite furnace to minimize the contribution of other materials to the resultant X-ray powder diffraction (XRPD) data pattern. The cylindrical sample was then inserted into a Cr-doped MgO octahedron with 10 mm-edge featuring two X-ray windows filled with boron epoxy rods. A detailed scheme of the assembly is available as supporting information, S2. The assembly was compressed using eight cubic WC anvils with a 5 mm truncation.

XRPD data were collected every 3 s on a PILATUS 3X CdTe 900K-W-ESRF 2D detector [details available in Crichton et al. (2023)] and a wavelength of 0.23262 Å for Run 5, 0.23393 Å for Run 6. Pressure and temperature conditions have been derived combining the press calibration (i.e. the relation between oil pressure and anvil pressure), and the equation of state of the sample and the calibrant. Pressure at high temperature has been obtained using the PVT equations of state of gold and platinum, respectively (Jamieson et al., 1982). Due to calibration, pressure uncertainty in these runs reflects an error approximately one order of magnitude greater than that of the single-crystal experiments, further compounded by temperature uncertainties. As a result, these data do not provide accurate absolute P–T conditions, but all measurements are affected by the same systematic error. Consequently, this dataset remains reliable for constraining the relative behavior and topology of phase boundaries. The paths followed during experiments are reported in Fig. 1(b), while the P–T conditions of the phase transitions observed are reported as supporting information, (S3 and S4). All XRPD patterns were automatically integrated at the beamline using detector geometry calibrated against a standard LaB₆ reference sample. Selected patterns have been refined with the Rietveld method using GSAS-EXPGUI and GSAS-II (Toby & Von Dreele, 2013) based on the structural models obtained from the single-crystal experiments reported in this study (refinements are available as supporting information, S5).

3. Results and discussion

3.1. High-pressure experiments and equation of state

The high-pressure experiments allowed the identification of new high-pressure polymorphs. Run 1 and Run 2 unit-cell parameters variation with pressure/temperature are reported in supporting information, S6 and S7, respectively. Experiment Run 1 showed two interesting pressure-induced phase transitions occurring at 1.33 (5) GPa and 11.46 (5) GPa. The two new structures, here referred to as ɛ and ɛ-II, are the result of second-order phase transitions, involving a slight re-arrangement of the structure and a variation in symmetry. This aspect will be discussed in detail in the next paragraph. Data collected during Run 1 [Fig. 2(a)] allowed the determination of the bulk modulus of the two new structures (Table 2). The same phase transition γ–ɛ has been observed in Run 2 [Fig. 2(b)], where data collected at low pressure allowed the determination of the bulk modulus of the γ phase. In experiment Run 2, silicone oil remains hydrostatic up to 4 GPa. In both experiments, volume–pressure data have been fitted using the EoSFit7_GUI program (Gonzalez-Platas et al., 2016) with a second-order Birch–Murnaghan equation of state (Birch, 1947). The obtained parameters and refinement details are given in Table 2.

Table 2
Equation of state parameters obtained by fit

Run	Phase	V0	K0	maxΔP	wχ^2
2	γ	273.4 (4)	41 (2)	−0.11	0.4
1	ɛ	549.7 (9)	47.6 (8)	0.20	2.52
1	ɛ-II	1597 (5)	59.8 (12)	1.40	26.20

Figure 2 Volume variation with pressure observed in (a) Run 1 and (b) Run 2. Where not displayed, e.s.d. error bars are smaller than symbols.

The second-order phase transition γ → ɛ does not have a significant effect on the compressibility of the structure, and the bulk modulus of the two polymorphs is not different within 3σ (Table 2). The compressibility of ɛ-II, in contrast, seems to be lower than the other structures, even though having only high-pressure data could result in an error on the bulk modulus estimation.

3.2. Crystal structures of ɛ and ɛ-II Na₂CO₃

As described by De Wolff & Tuinstra (1986), the sodium carbonate structure can be schematized as face-sharing columns of Na octahedra, parallel to the c axis, with Na and CO₃ groups in between the columns. This representation is particularly efficient in underlining the relationship between the different structural configurations that this phase can have. Fig. 3 shows the structure of the β polymorph along with the two new structures, referred to as ɛ and ɛ-II. In Run 1, we observed a direct phase transition from the incommensurate structure γ to the ɛ structure, and then from ɛ to ɛ-II. The ɛ structure has been observed in all the experiments performed. ɛ-II, observed above 11.46 (5) GPa, is a distinct polymorph characterized by a triplication of the b axis (supporting information, S8). Unit-cell parameters and space groups of the different polymorphs are given in Table 3. While the β structure shows all CO₃ groups in a co-planar distribution perpendicular to the c axis (Fig. 3), the ɛ structure can be described as the superimposition of two planar modules (i.e. A and B) with different orientation of CO₃ groups. The tilting of the CO₃ groups with respect to the c axis implies a deformation of the Na columns and a doubling of the unit-cell parameter c, as well as a reduction in symmetry from C2/m to Cc. The disorder in the CO₃ groups orientation is even more evident in the ɛ-II structure, where the tilting varies also horizontally within the same planar unit, implying a tripled b unit-cell parameter. The atomic positions and refinement details of the two new structures are available in the supporting information (S9 and S10).

Table 3
Unit-cell parameters and space groups of different Na₂CO₃ polymorphs

Data from this study are obtained by single-crystal XRD.

Phase	a (Å)	b (Å)	c (Å)	β (°)	Volume (Å^3)	Space group	P (GPa)	T (K)	Source
β	9.013 (2)	5.237 (2)	6.321 (2)	96.83 (2)	295.82	C2/m	0	673	de Pater (1979)
ɛ	8.832 (5)	5.2024 (5)	11.8155 (13)	102.03 (5)	531.0 (3)	Cc	1.92 (5)	300	This study, Run 1
ɛ-II	8.526 (3)	14.9693 (1)	11.0872 (1)	103.75 (4)	1374.5 (5)	Cc	12.32 (5)	300	This study, Run 1

Figure 3 Schematic representation of the β structures. In blue Na polyhedral, in red CO3 groups and in yellow Na atoms.

3.3. HP-HT experiments and phase diagram

SC-XRD has been performed in experiments Run 3 and Run 4. Specifically, Run 3 confirmed the stability of the ɛ phase even under high-temperature conditions, while Run 4, performed at relatively low-pressure conditions [i.e. up to 2.44 (5) GPa], allowed the identification of the γ–β–α phase transitions with temperature. These experiments served as accurate and precise anchor points in the building of the phase diagram. Variation of unit-cell parameters with pressure/temperature for Run 3 and Run 4 are given in S11 and S12 of the supporting information, respectively. In situ MA XRPD experiments (Run 5 and Run 6) allowed the identification of the same phase transitions at different pressure–temperature conditions (supporting information, S3 and S4). Fig. 4 shows the variation of the XRPD peaks position during P and T variations in Run 5.

Figure 4 XRPD peaks position variation in time in Run 5. Vertical axis shows the frame number, each frame has been collected for 3 s. P and T are estimated following the procedure described in Methods and are subject to significant uncertainty. The black areas represent filters applied to the MgO peak positions to enhance the visualization of the sample contribution.

The phase transitions and their reversibility observed in LVP experiments (Fig. 5) further constrained the topology of the phase boundaries.

Figure 5 Experimental points obtained in Run 5 and Run 6. We report an estimated error on temperature of ±50 K. Linear trendline of data is shown to assess the relative orientation of the phase boundaries.

4. Conclusions

We report two new high-pressure sodium carbonate structures, ɛ and ɛ-II. The ɛ polymorph is characterized by a doubling of the c unit-cell parameter compared to the γ and β polymorphs, and crystallizes in space group Cc. Its bulk modulus is 47.6 (8) GPa and is not significantly different from that of the γ phase. Its stability is confirmed by in situ single-crystal experiments and in situ LVP experiments. By exploring a range of pressure–temperature conditions, we observed multiple phase transitions and their reversibility, allowing us to constrain the topology of the Na₂CO₃ phase diagram. Using experimental single-crystal data as precise and accurate fixed points, and combining them with the sub-parallel phase-boundary topology observed in LVP experiments, we propose an updated phase diagram for sodium carbonate (Fig. 6). Interestingly, our data also align well with the phase boundary reported by Gavryushkin et al. (2019) for the hexagonal P6₃/mcm polymorph, stable above 12 GPa. The high-pressure stability limit of the ɛ phase observed in this study is consistent with the phase transition to the hexagonal polymorph reported by Gavryushkin et al. (2019). The ɛ-II structure, observed above 11.46 (5) GPa may represent a low-temperature polymorph or a metastable phase of Na₂CO₃.

Figure 6 Phase diagram of Na2CO3 obtained from experimental data. Symbols represent the different experimental runs as reported in the legend, colors represent the different polymorphs (α in blue, β in green, γ in black, ɛ in red).

Compared with the literature (Shatskiy et al., 2015), we observe a widening of the incommensurate γ-phase stability field toward higher pressures and temperatures. Both single-crystal and LVP experiments clearly show a positive P–T slope for the γ → β → α transition sequence.

Our results also show a wide stability field for the ɛ polymorph at pressures and temperatures compatible with the Earth's mantle. These results provide robust experimental benchmarks that improve the reliability of the phase diagram and clarify the relationships between existing and newly identified phases. The updated phase diagram offers a more complete framework for interpreting the structural evolution of sodium carbonate under pressure and temperature, and provides essential constraints for future experimental, computational and geological studies.

Supporting information

Computing details

(epsilon-Na2CO3_P5)

Crystal data

Na2CO3	β = 102.03 (5)°
Mr = 105.99	V = 531.0 (3) Å^3
Monoclinic, Cc	Z = 8
Hall symbol: C -2yc	F(000) = 416
a = 8.832 (5) Å	Dx = 2.652 Mg m^−3
b = 5.2024 (5) Å	T = 298 K
c = 11.8155 (13) Å

Data collection

ID15B beamline diffractometer	Rint = 0
Radiation source: synchrotron	θmax = 18.2°, θmin = 2.0°
411 measured reflections	h = −6→7
411 independent reflections	k = 0→7
358 reflections with I > 3σ(I)	l = −14→14

Refinement

Refinement on F	49 parameters
R[F^2 > 2σ(F^2)] = 0.051	0 restraints
wR(F^2) = 0.061	2 constraints
S = 4.09	Weighting scheme based on measured s.u.'s w = 1/(σ^2(F) + 0.0001F^2)
411 reflections	(Δ/σ)max = 0.035

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
Na1	0.4984 (13)	0.0404 (10)	0.2194 (4)	0.0235 (13)*
Na2	0.1604 (17)	0.0072 (10)	0.0907 (5)	0.0173 (5)*
Na4	0.6657 (15)	−0.4588 (4)	0.3422 (5)	0.0168 (5)*
O2	0.090 (2)	−0.3357 (17)	−0.0370 (8)	0.0221 (17)*
O3	0.294 (2)	−0.2601 (15)	0.2547 (6)	0.023 (2)*
O4	0.3827 (18)	0.0156 (13)	0.0081 (6)	0.0214 (18)*
Na3	0.8368 (15)	0.0374 (10)	0.4642 (4)	0.0266 (14)*
O5	0.539 (2)	0.2520 (13)	0.4331 (6)	0.0185 (18)*
O6	0.235 (2)	−0.6623 (18)	0.2153 (8)	0.027 (2)*
O7	0.4585 (17)	−0.4878 (13)	0.1796 (6)	0.0196 (18)*
C1	0.002 (2)	−0.5136 (16)	−0.0320 (7)	0.0090 (18)*
C2	0.329 (2)	−0.4625 (19)	0.2157 (7)	0.013 (2)*

(epsilonII-Na2CO3_P16)

Crystal data

Na2CO3	β = 103.75 (4)°
Mr = 318	V = 1374.5 (5) Å^3
Monoclinic, Cc	Z = 24
Hall symbol: C -2yc	F(000) = 1248
a = 8.526 (3) Å	Dx = 3.073 Mg m^−3
b = 14.9693 (1) Å	T = 298 K
c = 11.0872 (1) Å

Data collection

ID15B beamline diffractometer	Rint = 0
Radiation source: synchrotron	θmax = 18.4°, θmin = 2.2°
1188 measured reflections	h = −6→7
1186 independent reflections	k = 0→20
943 reflections with I > 3σ(I)	l = −14→14

Refinement

Refinement on F	145 parameters
R[F^2 > 2σ(F^2)] = 0.061	0 restraints
wR(F^2) = 0.068	2 constraints
S = 3.63	Weighting scheme based on measured s.u.'s w = 1/(σ^2(F) + 0.0001F^2)
1186 reflections	(Δ/σ)max = 0.027

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
Na1	0.9972 (15)	0.3547 (3)	0.9865 (6)	0.0177 (11)*
Na2	0.3158 (13)	0.3601 (4)	0.8588 (5)	0.0225 (12)*
Na3	0.8103 (13)	0.5085 (4)	0.8698 (5)	0.0249 (14)*
Na4	0.7999 (14)	0.1861 (3)	0.8708 (6)	0.0251 (14)*
Na5	0.4675 (14)	0.1783 (4)	0.2360 (5)	0.0235 (11)*
Na6	0.4614 (16)	0.1909 (3)	0.9854 (6)	0.0212 (12)*
C1	0.646 (3)	0.3474 (5)	0.8608 (10)	0.0015 (19)*
O1	0.548 (2)	0.2783 (6)	0.8584 (9)	0.025 (2)*
C2	0.653 (3)	0.6794 (6)	0.8664 (10)	0.005 (2)*
O2	0.840 (2)	0.5900 (5)	1.0721 (8)	0.0166 (18)*
O3	0.424 (2)	0.2978 (5)	0.1176 (9)	0.031 (2)*
O4	0.260 (2)	0.5127 (5)	0.8146 (7)	0.0196 (19)*
O5	0.2976 (19)	0.4224 (5)	1.0582 (7)	0.0118 (16)*
Na7	0.4613 (15)	0.5187 (3)	0.9917 (5)	0.0167 (11)*
O6	0.694 (2)	0.8215 (5)	1.1633 (8)	0.0194 (19)*
C3	0.145 (3)	0.5147 (6)	0.8576 (11)	0.008 (2)*
Na8	0.1646 (12)	0.1520 (3)	0.6029 (5)	0.0167 (12)*
O7	0.396 (2)	0.3906 (5)	0.6294 (8)	0.0171 (19)*
Na9	0.1328 (13)	0.4708 (3)	0.6142 (5)	0.0221 (12)*
O8	0.173 (2)	0.3156 (5)	0.6500 (8)	0.020 (2)*
Na10	0.4703 (15)	0.1639 (4)	0.7339 (6)	0.0202 (9)*
O9	0.261 (2)	0.1586 (5)	0.3236 (8)	0.018 (2)*
O10	0.186 (2)	0.0046 (5)	0.1539 (8)	0.024 (2)*
O11	0.146 (2)	0.2579 (5)	0.9076 (8)	0.022 (2)*
O12	−0.244 (2)	0.3442 (5)	0.3127 (8)	0.019 (2)*
Na11	0.4741 (12)	0.4839 (3)	0.2368 (5)	0.0168 (12)*
O13	0.358 (2)	0.2514 (5)	0.5650 (8)	0.0148 (18)*
O14	0.096 (2)	0.4152 (5)	0.4106 (8)	0.0189 (18)*
Na12	0.1306 (13)	0.1766 (3)	0.1065 (5)	0.0209 (12)*
O15	0.037 (2)	0.4484 (5)	0.8329 (9)	0.028 (2)*
O16	0.043 (2)	0.1232 (5)	0.8813 (9)	0.026 (2)*
O17	0.398 (2)	0.0529 (5)	0.6002 (8)	0.017 (2)*
O18	0.092 (2)	0.0766 (5)	0.3982 (8)	0.019 (2)*
C4	0.323 (4)	0.3192 (9)	0.6201 (17)	0.028 (3)*
C5	0.313 (4)	0.3454 (9)	0.1165 (15)	0.023 (3)*
C6	0.312 (4)	0.0125 (9)	0.1089 (16)	0.030 (4)*