Introduction

Fluorite is one of the best-known fluorescent minerals, and its specific properties led G. Stokes to name this phenomenon “fluorescence”. The luminescence of natural fluorite crystals was measured by various methods, among which the photoluminescence (PL) was most often used and significant. The luminescence of fluorite is connected mainly to the presence of rare-earth ions in it. The electrostatic stability of CaF2 lattice in presence of RE3+ ions demands the additional negative charge, mainly as an F− ion in interstitial sites. The effective symmetry around rare-earth ions could be O h (for Pr3+, Nd3+, Sm3+, Eu3+, Gd3+, Tb3+, Dy3+, Er3+ and Yb3+ ions), C 3v (for Gd3+, Tb3+, Dy3+, Ho3+, Er3+ and Yb3+ ions) or C 4v (for Ce3+, Nd3+, Gd3+, Tb3+, Dy3+, Ho3+, Er3+ and Yb3+ ions), what was confirmed by EPR measurements (Weber and Bierig 1964). There has been a large number of research papers on the luminescence of natural fluorite measured by the steady-state method (Illiev et al. 1988; Aierken et al. 2000, 2003; Bodył 2006; Petit et al. 2007; Czaja et al. 2008; Bodył 2009) or the time-resolved method (Gaft et al. 1998, 2001a, b, 2005, 2008). Moreover, the synthetic CaF2, BaF2 and SrF2 crystals doped with RE ions were investigated intensively (Wood and Kaiser 1962; Kirilyuk 1973; Fenn et al. 1973; Tallant and Wright 1975; Seelbinder and Wright 1979; Chrysochoos et al. 1982, 1983; Illiev et al. 1988; Caldino et al. 1989; Oskam et al. 2002). The splitting of ground and excited levels of Gd3+, Er3+ and Ho3+ ions were determined, and luminescence lines were assigned to the C 4v and C 3v symmetry sites on the basis of EPR and electronic optical measurements (Rector et al. 1966; Tallant and Wright 1975; Seelbinder and Wright 1979; Mujaji et al. 1992). In the case of heavy-doped CaF2 synthetic crystals and their analogues, other luminescence centers, such as pairs or clusters, have been discussed by researchers, too (Fenn et al. 1973; Seelbinder and Wright 1979).

In this paper, we have demonstrated the possibility of steady-state photoluminescence measurement for the purpose of identification of 4f ions using the mutual relationship between excitation and emission spectra. For the first time ever, the emission and excitation spectra of Pr3+, Nd3+, Tm3+ and Yb3+ in natural fluorite crystal were investigated by this method. Moreover, we have proposed the most effective excitation for holmium ion emission. Furthermore, the tentative assignment emission line to the respective transition in C 3v and C 4v symmetry sites was made from the time-resolved spectra of Er3+ and Ho3+ ions.

Sample properties and measurement conditions

The fluorite crystals studied in this work were found in a lens-like biotite-hornblendite pegmatite in Paszowice (Sudety Mountains, Poland). These crystals were transparent and pale yellow, but rather small (1 mm3). The crystal phase was confirmed by X-ray diffraction (Philips PW 3710). The amount of lanthanides was measured by the ICP-MS method in ACME Laboratory (Canada). The marked europium anomaly (0.004) and the plot of REE (rare-earth element) content normalized to chondrite C1 (Fig. 1) are typical for pegmatite fluorite. The ΣREE = 15,765.68 ppm, ΣΗREE = 14,588.79 ppm and ΣΗREE/ΣLREE = 12.4, where LREE–light rare-earth element, i.e., La–Eu, HREE–hard rare-earth element), i.e., Gd–Lu and Y). However, when the Y is excluded, the contents of light, medium and heavy REE are similar.

Fig. 1
figure 1

REE content in fluorite (Paszowice, Poland) normalized to chondrite C1

Full size image

To find the most convenient conditions to measure Ho3+ luminescence in the presence of Er3+ ions, the luminescence spectra of phosphate glass doped by holmium and erbium: 44 % P2O5 + 25 % CaO + 15 % BaO + 15 % SrO + 1 % Ho2O3, 44 % P2O5 + 25 % CaO + 15 % BaO + 15 % SrO + 1 % Er2O3 were synthesized.

Phosphate glasses doped with lanthanide ion or ions can be easily synthesized–easier than CaF2 crystals. However, the optical properties of 4f ions in these glasses, especially Ωi parameters, are different from those in other host materials, although the positions of emission and excitation bands can be treated as a standard (with accuracy measured in ± nm) in order to identify each particular 4f ion. We have observed that it is especially useful for Pr3+–Sm3+ ions in apatite crystals (Bodył et al. 2009) and for Er3+–Ho3+ ions in fluorite and scheelite crystals (Czaja et al. 2008). To enhance the emission of Ho3+ ions in the presence of Er3+ ions in the spectral range of 550 nm, a 415 nm excitation was chosen. For this excitation, the transition 5I8 → 5G5 of Ho3+ ions can be observed besides the transition 4I15/2 → 2H9/2 of Er3+ ions. However, the emission from the 2H9/2 level was not measured, in contrast to the emission from 4S3/2.

The steady-state fluorescence measurements for Ce3+, Pr3+, Sm3+, Eu2+, Tb3+, Dy3+, Ho3+, Er3+ and Tm3+ were performed using a Jobin–Yvon (SPEX) spectrofluorimeter FLUORLOG 3-12 at room and low temperatures using a 450 W xenon lamp, a double-grating monochromator, and a Hamamatsu 928 photomultiplier. Other steady-time measurements were done for Nd3+, Sm2+, Er3+ and Yb3+ using an Edinburgh Instruments FLS920 spectrofluorimeter with a xenon lamp and Hamamatsu 928 or Hamamatsu R5509-72 photomultipliers or Physik LPD3000 laser (pumped by a Lambda Physik LPX100 excimer laser). The time-resolved emission spectra and decay times were measured using a GDM–1000 double-grating monochromator, equipped with a Hamamatsu 928 photomultiplier. The resulting luminescence signal was stored in a Stanford model SRS 250 Boxcar Integrator coupled with a PC computer. The emission-line accuracy was 0.1 nm. Because the sensitivity of the Hamamatsu 928 or Hamamatsu R5509-72 photomultipliers is almost constant in the range 400–700 nm, no correction curve was needed.

The luminescence decay curves were excited by applying a short impulsive light from an OPO Optical Parametrical Oscillator pumped by the third-harmonic of a YAG:Nd laser. The decay kinetics of excited states were recorded utilizing a Tektronix model TDS 3052 digital oscilloscope. The decay time and gate width were chosen according to the decay time of each expected luminescence center.

Results and discussion

Steady-state measurements for 4f-5d transitions

The luminescence spectrum of Ce3+ ion of the studied fluorite was very similar to those which were known from Aierken et al. (2000) or Bodył (2009). For λexc. = 302 nm, the intensive bands at 320 and 343 (336) nm, called by Aierken et al. (2000) A and B, were measured and could be assigned to the transitions from 5d(Eg) → 4f(2F5/2) and 4f(2F7/2), respectively. The Eu2+ ions usually caused a very characteristic blue emission, which was usually measured at 420 nm as the electronic transition 4f 6 5d(Eg) → 4f 7(8S7/2). However, for this crystal, the said emission was not measured, due to a very low Eu concentration (Table 1).

Table 1 Abundance (ppm) of rare-earth impurities in natural fluorite from Paszowice, compared with Aierken et al. (2003)
Full size table

The emission of Sm2+ ions was measured at low temperatures (Fig. 2). In fluorite, the excited 4f 6 level (5D0) is just below the lowest excited 4f 5 5d(Eg) level (Wood and Kaiser 1962), so the energy of the lowest level of te 4f 5–5d 1 excited electronic configuration could interact significantly with the 5Dj levels of the 4f ground configuration and have a strong influence on the optical properties of Sm2+. The sharp and intensive emission line at 683 nm (14,641 cm−1) could be attributed to the zero-phonon line (ZPL) of the T1u(4f 5 5d) → A1g(7F0(4f 6)) transition. Other emission lines at the longer wavelength part of the spectrum, that is, 693 nm (14,430 cm−1), 697 nm (14,347 cm−1) and 699 nm (14,306 cm−1), have a vibronic origin according to the Elcombe and Pryor (1970) and correspond to the ZPL minus the frequency of modes 210, 250 and 355 cm−1, respectively. The emission lines at 708, 720, 730, 748, 761 and 793 nm are assigned to the 5D0 → 7FJ (J = 0,1,…,6), and the 365, 452, 467, 485 and 494 nm lines on the excitation spectrum to the 4f 6(7F0) → 4f 5(6Fj and 6Hj)5d(T2g) transitions, respectively.

Fig. 2
figure 2

Luminescence spectra of Sm2+: a excitation spectrum monitored at λem = 682 nm at T = 6 K (black line), b and c emission spectra measured at λexc. = 485 nm at T = 100 K and T = 6 K, as red and blue lines, respectively

Full size image

Steady-state measurements for 4f-4f transitions

Neodymium (Nd3+) is one of the most efficient RE centers in minerals, and its characteristic spin-allowed transitions in NIR range are 4F3/2 → 4I9/2 (864 and 895 nm) and 4F3/2 → 4I11/2 (1,065 nm). The NIR emission spectra of Nd3+ ion in minerals were measured so far by the time-resolved technique. The steady-state luminescence spectrum of Nd3+ ions in our fluorite crystal was measured and presented on Fig. 3. The decay time of the 4F3/2 state of Nd3+ ions in this fluorite had the value of τ = 635 μs. A similarly long decay time of this transition was earlier found in Cs2NaNdCl6 [ρ(Nd) = 3.2 × 1021 cm−3] and in Cs2NaNd0.01Y0.99 Cl6 crystals (Tofield and Weber 1974) and it amounted to 1.23 and 4.1 ms, respectively. This means that Nd3+ ions in our fluorite crystal occupy the O h symmetry site because only strict octahedral coordination of Nd3+ discourages electric-dipole electronic transitions. Subramanian and Mukherjee (1987) predicated that Nd3+ ions in fluorite could occupy the O h and C 4v sites.

Fig. 3
figure 3

Luminescence emission spectrum of Nd3+, Yb3+ and Er3+ measured for λexc. = 521 nm, measured at T = 300 K

Full size image

Moreover, additional emission peaks of Er3+ at 1,529 nm (4I13/2 → 4I15/2) and Er3+ (4I11/2 → 4I15/2) together with Yb3 (4F5/2 → 2F7/2) at 978 nm were observed. The luminescence lifetime for 4F5/2 state of Yb3+ ion in this fluorite was measured and was equal to τ = 9.3 ms, which was coherent with the results obtained earlier. For the Yb3+ ion in C 4v site symmetry in synthetic CaF2, Petit et al. (2007) have observed that the luminescence lifetime was τ = 8 ms.

For the Pr3+ ion, the characteristic emission lines and corresponding transitions are 480–500 nm (3P0 → 3H4), 650–670 nm (3P0 → 3F2), 750–770 nm (3P0 → 3F4), 610–630 nm (1D2 → 3H4) and (3P0 → 3H6), 400–410 nm (1S0 → 1I6). Likewise, some other transitions (5d-4f) were measured in the UV region (Gaft et al. 2005). The steady-state luminescence measurements of Pr3+ in minerals are difficult because radiative transitions of this ion are hidden by the stronger emission of Sm3+ ion (600–650 nm), Dy3+ ion (470–490 nm) or Nd3+ in the IR range (870–900 nm) (Gaft et al. 2005). It has been noticed (Bodył et al. 2009) that the emission of the Pr3+ ion is most intensive for λexc. = 442 nm. For such a condition (Fig. 4), the diagnostic emission lines connected to the 3P0 → 3H6 and 3P0 → 3F2 transitions, appeared at 614 and 641 nm, respectively. Other emission lines on this spectrum (598 and 607 nm) could be assigned to the 4G5/2 → 6H7/2 transition of the Sm3+ ion. Chrysochoos et al. (1982, 1983) revealed that Pr3+ ions occupied mainly O h and C 4v positions.

Fig. 4
figure 4

Luminescence spectra of Pr3+: a and b excitation spectra monitored at 614 and 641 nm, as red and blue lines, respectively, c emission spectrum measured for λexc. = 442 nm, measured at T = 300 K (black line)

Full size image

The most efficient Sm3+ luminescence is usually measured for λexc. = 399–401 nm (Bodył et al. 2009). The luminescence spectra of our fluorite crystal (Fig. 5) have seemed similar to those presented by Aierken et al. (2003). Besides the emission of Sm3+: 560 nm (4G5/2 → 6H5/2), 599 and 606 nm (4G5/2 → 6H7/2) and 644 and 653 nm (4G5/2 → 6H9/2), emission lines of Dy3+ at 574 nm (4F9/2 → 6H13/2), Er3+ at 538 nm (4S3/2 → 4I15/2) and at 653 nm (4F9/2 → 4I15/2), and Sm2+ at 683 nm (4f 5 5d → 4f 6) were observed as well.

Fig. 5
figure 5

Luminescence spectra of Sm3+: a and b excitation spectra monitored at λem. = 596 and 606 nm, as red and blue lines, respectively, c emission spectrum measured for λexc. = 399 nm, measured at T = 300 K (black line)

Full size image

It was established that the proper λexc. for Dy3+ ion is equal 348 nm. The emission lines at 478 nm (4F9/2 → 6H15/2) and 574 nm (4F9/2 → 6H13/2) were measured (Fig. 6). When other excitation was used, the emission of another ion (or ions) was measured. For example, the emission of erbium ion (542 nm) besides dysprosium ion was measured for λexc. = 323, 362 or 387 nm, while an emission of Sm3+ and Pr3+ ions was observed for λexc. = 449 nm. On the other hand, for λexc. = 348 nm, Aierken et al. (2003) have shown the emissions of Tb3+ as peaks at 495, 546, 582 and 623 nm. The first three of them were present on our fluorite spectrum (Fig. 6); however, no characteristic excitation lines of the Tm3+ ion (302, 317, 340, 350, 369, 376, 483 nm) were found on excitation spectra. Instead, for λem. = 495 and 582 nm, the Dy3+ lines (323, 348, 362, 387 nm) appeared, and for λem. = 546 nm, the excitation lines characteristic for the Er3+ ion were observed. It is likely that intensive Er3+ emission and excitation lines hide the terbium luminescence. Moreover, the complex nature of the 478 nm emission line of Dy3+ ion could be seen for other crystals (Bodył-Gajowska 2010).

Fig. 6
figure 6

Luminescence spectra of Dy3+: a and b excitation spectra monitored at λem. = 574 and 478 nm, as red and blue lines, respectively, c emission spectrum for λexc. = 348 nm, measured at T = 300 K (black line)

Full size image

The Tm3+ luminescence for minerals was measured mainly using the time-resolved method (Gaft et al. 2005) or by the cathodoluminescence (CL) technique. In this study, we demonstrate the evident emission of thulium ion (Fig. 7); when λexc. = 353 nm, the transitions 1D2 → 3H4 of Tm3+ at 453 nm as well as 4F9/2 → 6H15/2 (478 nm) and 574 (4F9/2 → 6H13/2) of Dy3+ ion and 4S3/2 → 4I15/2 (538 nm) of Er3+ ion were observed. The concentration of thulium in our fluorite crystal was rather high, but lower than in the crystal of Aierken et al. (2003), so the obtained spectrum could be deemed a success.

Fig. 7
figure 7

Luminescence spectra of Tm3+: a excitation spectrum monitored at λem. = 453 nm, b emission spectrum measured at λexc. = 353 nm, measured at T = 300 K

Full size image

The very intensive luminescence of Er3+ ion in the VIS and NIR part of spectrum is well known. Under λexc. = 377 nm, strong several lines appear (Fig. 8) and are connected to following electron transitions: 523 nm to 2H11/2 → 4I15/2 and 538–554 nm to the 4S3/2 → 4I15/2. The NIR emission spectrum of Er3+ was measured at λexc. = 521 nm as 978 (974) nm and 1,529 nm lines, that is, 4I11/2 → 4I15/2 and 4I13/2 → 4I15/2 transitions, respectively (Fig. 3). The luminescence lifetime for the excited state of the 4S3/2 of Er3+ ion in this fluorite was measured (τ = 436 μs) and was close to that predicted by the Judd–Ofelt analysis for 4S3/2 → 4I15/2 transition of Er3+ ion in phosphate glass (Mazurak et al. 2010). When λexc. = 415 nm is used, emission line 523 nm of the erbium ion disappears almost completely, and the most intensive lines become 536 nm and 552 nm. With excitation at 415 nm (24,096 cm−1), the 2H11/2 at 523 nm (19,120 cm−1) is some 5,000 cm−1 lower in energy and this gap is too large for a phonon-assisted decay. However, the Er3+ levels nearest to 415 nm are 4F3/2 and 4F5/2 at ~22,300 cm−1 (448 nm), and these can be easily populated as there are few phonons below the pump. It is possible to populate the 4S3/2 level via energy exchange mechanisms even for weakly coupled Er3+ ions. As a result, an efficient emission is observed from 4S3/2, not 2H11/2.

Fig. 8
figure 8

Luminescence spectra of Er3+: a and b excitation spectra monitored at λem. = 539 and 523 nm (red and blue lines, respectively), c and d emission spectra measured for λexc. = 415 nm and for λexc. = 377 nm, (green and black lines, respectively), all measured at T = 300 K

Full size image

The comparison of the emission spectrum of fluorite with spectra of phosphate glasses doped with Er3+ and Ho3+ (Fig. 9) allows us to conclude that λexc. = 415 nm evidently enhances the green luminescence of Ho3+ ions, that is, the 5S2 → 5I8 transition. Aierken et al. (2003) have indicated that the emission line at 553 nm is also characteristic for the 5S2 → 5I8 transition in the Ho3+ ion.

Fig. 9
figure 9

Excitation spectra of a Ho3+ doped phosphate glass (green line) and b Er3+ doped phosphate glass (red line) monitored λem. = 552 nm, c of fluorite Paszowice monitored λem. = 550 nm black line), measured at T = 300 K

Full size image

The comparison of results for steady-state measurements was put together in Table 2.

Table 2 The most convenient conditions for steady-state measurements of some RE ions; excitation and emission lines and electronic transitions
Full size table

Time-resolved measurements

In order to properly identify the origin of emission lines in the 539–554 nm part of the spectrum, to assign them to Er3+ or Ho3+ ions, and also to determine the crystal sites occupied by these ions in fluorite lattice, time-resolved measurements were performed. The lifetimes for Ho3+ 540 nm and Er3+ 545 nm emissions measured by Gaft et al. (2001a, b) are similar and equal 5 μs and 23 μs, respectively. When the delay time is 10 μs and the gate width 10 μs, the emission of Ho3+ will be observed as well, while the Er3+ center is still in the excited state during the time of measurement and does not participate in the emission. By contrast, for delay time 10 μs and gate width 30 μs, the emission of Ho3+ will be already quenched, while the emission of Er3+ will remain.

On the time-resolved emission spectra, it is normal for the most intensive Stark multiplets to appear. The crystal field can split the energy levels of the RE ion and remove their degeneracy, as the complex character of absorption and emission spectra have often revealed. The number of multiplets, commonly referred to as Stark levels, depends on the quantum number J of 2S+1LJ terms and on the crystal site symmetry of the ion. The emission lines belong to transitions between Stark’s multiplets of excited and ground levels. The ground level is usually named Z, and the excited levels are designated as “Y, A, B, D, E”, etc. According to these designations, the 5F4,5S2 levels of Ho3+ and 5S3/2 level of Er3+ are designated as E. The degeneracy of the ground level of Ho3+ (5I8) and of excited 5F4,5S2 levels for the tetragonal symmetry site C 4v amount to 13 and 11, while for the trigonal symmetry site C 3v –11 and 9. The degeneracy of the excited level for the Er3+ ion is double and of ground level it is eightfold, for both symmetry sites.

For low-temperature measurements, the separation of holmium from erbium emission lines was perfectly visible. These lines are shown on Fig. 10a, b for holmium (5F4,5S2 → 5I8) and erbium4S3/2 → 4I15/2 transitions, respectively. All of these transitions are numbered, counted in [cm−1] and shown in Table 3. The emission lines measured for the fluorite from Paszowice are different from those measured for synthetic crystals by Rector et al. (1966), Dieke (1968), Tallant and Wright (1975), Seelbinder and Wright (1979) or Mujaji et al. (1992). After Seelbinder and Wright (1979) or Mujaji et al. (1992), the most intensive emission lines of the Ho3+ ion in the C 4v symmetry site were measured at 18,606, 18,490 and 18,448 cm−1 and identified as E1–Z1, E1–Z4 and E1–Z6 transitions, respectively, while for the C 3v symmetry site, the most intensive emission lines and transitions were 18,566 cm−1 (E1–Z1) cm−1, 18,539 cm−1 (E1–Z2), and 18,501 cm−1 (E1–Z3). According to Rector et al. (1966) and Tallant and Wright (1975), the most intensive emission lines of Er3+ ion in C 4v symmetry site were measured at 18,622, 18,601, 18,539, 18,518 and 18,136 cm−1 and identified as E2–Z1, E2–Z2, E1–Z1, E1–Z2 and E1–Z6 transitions, respectively, while for the C 3v symmetry site, the most intensive emission lines and transitions were at 18,607 cm−1 (E2–Z1) cm−1, 18,591 cm−1 (E1–Z1), 18,547 cm−1 (E1–Z2), 18,364 cm−1 (E1–Z3) and 18,319 cm−1 (E1–Z5).

Fig. 10
figure 10

Time-resolved luminescence spectra of fluorite from Paszowice measured at λexc. = 415 nm; a emission lines of Ho3+: delay time = 10 μs, gate time = 10 μs (green line), b emission lines of Er3+: delay time = 10 μs, gate time = 30 μs (red line), measured at 10 K

Full size image
Table 3 The energy of transitions between Stark’s levels–excited and ground–and assignment to C 4v and C 3v symmetry sites
Full size table

There were some difficulties with assigning numerous emission lines to luminescence transitions and comparing them to the values of energy, which are known for synthetic CaF2:Ho3+ or CaF2:Er3+crystals. Furthermore, some lines are present on the emission spectrum for holmium ion (Fig. 10a) at the shorter wavelength, whose energy should correspond to the transitions from the higher energy sublevels of 5F4,5S2 excited level (E11-9) to the ground 5I8 level (Z1-4), according to Mujaji et al. (1992). Because the population of (E11-9) states at T = 10 K is very small, this assumption should be rejected. The above discrepancies have led us to accept the assumption that the energies of excited states of Ho3+ and Er3+ ions in natural crystals are different—insensibly but measurably—from the energies of excited levels in synthetic crystals. The reason for these differences is the different value of local crystal field strength, due to the more complicated chemical composition of a natural fluorite crystal as compared to its synthetic counterpart.

For the Ho3+ ion in the fluorite from Paszowice, we have assumed that:

  • The excited levels 5F4 and 5S2 are energetically separated, and besides the emission from the 5F4 level, the emission from the 5S2 level was measured as well, as was shown for a few other crystals: YSGG (Pugh et al. 1997), KGd(WO4)2 (Pujol et al. 2001), GdLiF4, YLiF4 and LuLiF4 (Walsh et al. 2005), YAG (Walsh et al. 2006), YAB (Baraldi et al. 2007) or YGG (Gruber et al. 2009). The 5F4 and 5S4 levels were usually separated in them by an interval of 210–300 cm−1;

  • The splitting of ground 5I8 level (ΔZ) and excited level 5S2 (ΔE) is different in this case than for a synthetic CaF2:Ho3+ crystal;

  • Similarly to a synthetic CaF2:Ho3+ crystal, the most intensive transitions for a holmium ion in a tetragonal site should be E1–Z1 and E1–Z4, but for a trigonal site, they could also be E1–Z1 and E1–Z3.

For the Er3+ ion in the fluorite from Paszowice, we have assumed that:

  • There are differences in the splitting of the ground 4I15/2 level (ΔZ) and the excited level 5S3/2 (ΔE) for our fluorite and synthetic CaF2:Er3+ crystals;

  • the energy of excited level 5S3/2 is different (lower) in comparison with a synthetic crystal;

  • Similarly to a synthetic CaF2:Er3+ crystal, the most intensive transition for erbium ion in a tetragonal site should be E2–Z1, E1–Z1, E1–Z2 and E1–Z5,6, but for trigonal site also E1–Z1, E1–Z2 and E1–Z3;

  • For the Er3+ ion in a CaF2 crystal, the energy of the Z4 level could not be identified.

The above assumptions allow us to make a tentative assignment of luminescence lines to a particular transition of holmium and erbium ions in C 3v and C 4v symmetry sites (Fig. 10; Table 3). The excited level 5S3/2 of Er3+ has lower energy by about 320 and 240 cm−1 for C 4v and C 3v sites, respectively, whereas the excited level 5S2 of Ho3+ has higher energy by about 114 and 37 cm−1 for C 4v and C 3v sites, respectively. The splitting of ground level (4I15/2) of the Er3+ ion occupying C 4v and C 3v symmetry sites in synthetic CaF2 is equal to 84 and 16 cm−1, respectively, while the splitting of the excited level 5S3/2 equals 452 and 461 cm−1. For the Ho3+ ion in synthetic CaF2, it was found that the splitting of the ground level is equal to 512 and 423 cm−1 for the C 4v and C 3v symmetry sites, respectively, while the splitting of the excited level equals 289 and 206 cm−1. From our tentative assignment, the splitting of the excited level for both ions seemed to be generally lower than in a synthetic crystal.

Summary

This study has shown that steady-state measurements can identify the 4f ions in natural crystals. The luminescence of Pr3+, Nd3+ and Tm3+ ions in fluorite crystal have been received by steady-state measurements; the luminescence of Ce3+, Sm2+, Eu2+, Sm3+, Dy3+, Er3+ and Yb3+ was measured as well. In comparison with earlier studies on this subject (e.g. Aierken et al. 2003), the intensive luminescence was measured for concentrations of RE ions 1.5–4.0 times smaller. The fluorite crystal studied in this paper is another example of a fluorite for which the violet luminescence did not dominate. Transitions between energy levels of the Ho3+ and Er3+ ions for a natural fluorite crystal were measured. The discrepancies between the energy of Stark levels of holmium and erbium ions in synthetic and natural crystals could have been caused by differences in crystal field strength of natural and synthetic crystals. In order to remove any doubt regarding the assignment of the emission transitions and to calculate the energy of Stark levels, it is necessary to measure the energy of transitions from other excited levels of Er3+ and Ho3+ to ground state and will be the subject of a separate study.