IChO 2018 · Theory problem 5

Bohemian garnet

The dodecahedral splitting diagram referred to in 5.4 and the formula defining \(x,y\) in 5.13 are not printed in this version of the problem.

PROBLEM 5

Bohemian garnet

Bohemian garnet (pyrope) is a famous Czech blood coloured semi-precious stone. The chemical composition of natural garnets is expressed by the general stoichiometric formula of \(\ce{A3B2(SiO4)3}\), where \(\ce{A^{II}}\) is a divalent cation and \(\ce{B^{III}}\) is a trivalent cation. Garnets have a cubic unit cell that contains 8 formula units. The structure comprises 3 types of polyhedra: the \(\ce{A^{II}}\) cation occupies a dodecahedral position (it is surrounded with eight O atoms), the \(\ce{B^{III}}\) cation occupies an octahedral position (it is surrounded with six O atoms) and \(\ce{Si^{IV}}\) is surrounded with four O atoms arranged into a tetrahedron.

The most common garnet mineral is almandine with the formula of \(\ce{Fe3Al2(SiO4)3}\). Its unit cell parameter is \(a = 11.50\ \text{Å}\).

5.1 Calculate the theoretical density of almandine.

The Bohemian garnet has the composition of \(\ce{Mg3Al2(SiO4)3}\). Pure compound is colourless and the colour of natural garnets comes from chromophores – transition metal cations that substitute the host material cations. The red colour of the Bohemian garnet comes from trace amounts of \(\ce{Cr^{III}}\) ions in the octahedral sites and \(\ce{Fe^{II}}\) ions in the dodecahedral sites.

5.2 Draw the splitting diagram for the \(\ce{[Cr^{III}O6]}^{\mathrm{oct}}\) d-orbitals and fill it with electrons.

5.3 Identify the 1st transition row element(s) whose trivalent cation(s) \(\ce{M^{III}}\) placed in an octahedral position is/are diamagnetic in the low-spin arrangement and paramagnetic in the high-spin arrangement.

5.4 The figure below shows d-orbitals splitting in the dodecahedral crystal field. Fill in the electrons for the \(\ce{[Fe^{II}O8]}^{\mathrm{dod}}\) chromophore for both existing arrangements.

5.5 Derive the inequalities (e.g. \(P < E_1 + E_2 + E_3\)) for the pairing energy (\(P\)) magnitude in relation to energies \(E_1\), \(E_2\) and \(E_3\) for both arrangements.

5.6 Assuming that \(P > E_3\), identify the 1st transition row element(s) whose divalent cation \(\ce{M^{II}}\) placed in dodecahedral position is diamagnetic in the low-spin arrangement and paramagnetic in the high-spin arrangement.

The figures below show simplified absorption spectra of four coloured minerals – blood-coloured Bohemian garnet, green uvarovite, blue sapphire and yellow-orange citrine.

Figure 1

5.7 Match the spectra with the minerals.

5.8 If illuminated with monochromatic blue-green light, how will the Bohemian garnet look?

☐ Red  ☐ Blue  ☐ Yellow-orange  ☐ Black
☐ Yellow ☐ Blue-green ☐ Violet  ☐ White

Andradite is another garnet mineral; its chemical composition is \(\ce{Ca3Fe2(SiO4)3}\). A double cation substitution – \(\ce{Ti^{IV}}\) for \(\ce{Fe^{III}}\) in the octahedral position and \(\ce{Fe^{III}}\) for \(\ce{Si^{IV}}\) in the tetrahedral position – gives rise to black schorlomite. Its chemical composition can be expressed as \(\ce{Ca3[Fe,Ti]2^{oct}([Si,Fe]^{tet}O4)3}\).

5.9 Calculate the percentage of \(\ce{Si^{IV}}\) ions in a sample of schorlomite that must be substituted with \(\ce{Fe^{III}}\), if we know that 5% of \(\ce{Fe^{III}}\) ions in octahedral position are substituted with \(\ce{Ti^{IV}}\).

The colour of the mineral is caused by two chromophores: \(\ce{[Fe^{III}O6]}^{\mathrm{oct}}\) and \(\ce{[Fe^{III}O4]}^{\mathrm{tet}}\). The central ions of both chromophores have equal number of unpaired electrons.

5.10 Draw the d-orbitals splitting diagrams for both chromophores and fill in the electrons.

Tetrahedral field causes a smaller splitting than the octahedral field \(\left(\Delta_{\mathrm{tet}} = \frac49\Delta_{\mathrm{oct}}\right)\). Surprisingly for the \(\ce{Fe^{III}}\) ion, the energy of the first d–d transition (although very weak) for the octahedral chromophore is smaller (\(11\,000\ \mathrm{cm^{-1}}\)) than for the tetrahedral one (\(22\,000\ \mathrm{cm^{-1}}\)).

5.11 Calculate the size of pairing energy (\(P\)) and the sizes of \(\Delta_{\mathrm{oct}}\) and \(\Delta_{\mathrm{tet}}\) splitting. Assume that the pairing energy is equal in both chromophores.

Synthetic garnet YAG (Yttrium Aluminium Garnet), used in optoelectronics, has the composition of \(\ce{Y3Al5O12}\). Its structure is derived from the general garnet structure \(\ce{A3B2(SiO4)3}\) by placing the ions \(\ce{Y^{III}}\) and \(\ce{Al^{III}}\) to the A, B and Si positions.

5.12 Based on your knowledge of the relative ion radii, determine which cation occupies which position.

5.13 For the use in LED technology, YAG is doped with \(\ce{Ce^{III}}\). Determine the values of \(x\) and \(y\) in the formula of YAG in which 5% of yttrium atoms are substituted with cerium. (If you don’t get result, use \(x = 2.25\) and \(y = 0.75\).)

5.14 The \(\ce{Ce^{III}}\)-doped YAG is prepared by annealing the mixture of \(\ce{Y2O3}\), \(\ce{Al2O3}\) and \(\ce{CeO2}\) in \(\ce{H2}\) atmosphere. Use the formula from 5.13, write down a balanced equation for this reaction with the smallest whole-number stoichiometric coefficients.

Doping the YAG structure with rare-earths ions enables the production of lasers with emission wavelengths ranging from the UV to the mid-IR region. In the scheme below, simplified f–f energy transitions of selected rare-earths ions are shown.

Figure 2

5.15 Which cation has a transition which corresponds to blue light emission.

☐ \(\ce{Er^{3+}}\)  ☐ \(\ce{Sm^{3+}}\)  ☐ \(\ce{Tm^{3+}}\)  ☐ \(\ce{Pr^{3+}}\)
☐ \(\ce{Yb^{3+}}\)  ☐ \(\ce{Nd^{3+}}\)  ☐ \(\ce{Tb^{3+}}\)

5.16 Calculate the emission wavelength of this light.

5.17 According to a legend, Noah used a stick with a garnet stone for illumination during his voyage. Assuming only the photoluminescence effect, determine the colour of the laser light emitted from his stick if the stone were the blood-coloured Bohemian garnet.

☐ Red  ☐ Blue  ☐ Yellow-orange  ☐ Black
☐ Yellow ☐ Blue-green ☐ Violet  ☐ White

Official English text of the problem.

Answers

For 5.4–5.6, use the usual dodecahedral level order \[ \varepsilon_a<\varepsilon_b<\varepsilon_c=\varepsilon_d<\varepsilon_e, \] with adjacent gaps \(g_L,g_M,g_H\), from bottom to top. The missing diagram prevents confirmation of their correspondence to \(E_1,E_2,E_3\).

  • 5.1: \(\boxed{\rho=4.348\ \mathrm{g\,cm^{-3}}}\).
  • 5.2: Chromium(III), \(\ce{Cr^{3+}}\), [Cr+3]: \(d^3\), configuration \(\boxed{t_{2g}^{3}e_g^{0}}\), with three unpaired electrons.
  • 5.3: Cobalt, through cobalt(III), \(\ce{Co^{3+}}\), [Co+3]: low-spin \(t_{2g}^{6}\), high-spin \(t_{2g}^{4}e_g^{2}\).
  • 5.4: Iron(II), \(\ce{Fe^{2+}}\), [Fe+2], is \(d^6\). Lower-spin: \(\boxed{a^2b^2c^1d^1e^0}\), two unpaired electrons. High-spin: \(\boxed{a^2b^1c^1d^1e^1}\), four unpaired electrons.
  • 5.5: Lower-spin when \(\boxed{P<g_M+g_H}\); high-spin when \(\boxed{P>g_M+g_H}\). If \(E_1,E_2,E_3\) run from bottom to top, the threshold is \(E_2+E_3\); if they run from top to bottom, it is \(E_1+E_2\).
  • 5.6: The possible diamagnetic lower-spin ions are titanium(II), [Ti+2], chromium(II), [Cr+2], and nickel(II), [Ni+2], requiring \(P<g_L\), \(P<g_M\), and \(P<g_H\), respectively. The condition \(P>E_3\) excludes the ion associated with that gap. For bottom-to-top \(E_1,E_2,E_3\), the answer is titanium and chromium; for top-to-bottom numbering, it is chromium and nickel.
  • 5.7: \(\boxed{\text{A: citrine; B: Bohemian garnet; C: uvarovite; D: sapphire}}\).
  • 5.8: \(\boxed{\text{Black}}\), in the intended absorption model.
  • 5.9: \(\boxed{3.33\%}\) of the silicon(IV) sites.
  • 5.10: Iron(III), \(\ce{Fe^{3+}}\), [Fe+3], is high-spin \(d^5\) in both sites: octahedral \(\boxed{t_{2g}^{3}e_g^{2}}\), tetrahedral \(\boxed{e^{2}t_2^{3}}\). Both have five unpaired electrons.
  • 5.11: In spectroscopic units, \[ \boxed{\frac{P}{hc}=30\,800\ \mathrm{cm^{-1}},\quad \frac{\Delta_{\mathrm{oct}}}{hc}=19\,800\ \mathrm{cm^{-1}},\quad \frac{\Delta_{\mathrm{tet}}}{hc}=8\,800\ \mathrm{cm^{-1}}}. \]
  • 5.12: Yttrium(III), [Y+3], occupies the eight-coordinate A sites. Aluminium(III), [Al+3], occupies both the six-coordinate B sites and the four-coordinate Si sites.
  • 5.13: Taking the formula as \(\ce{Y_xCe_yAl5O12}\), \(\boxed{x=2.85,\ y=0.15}\). The substituent is cerium(III), [Ce+3].
  • 5.14: \[ \boxed{\ce{57Y2O3 + 100Al2O3 + 6CeO2 + 3H2 -> 40Y2.85Ce0.15Al5O12 + 3H2O}}. \] The molecular reactant and product are hydrogen [H][H] and water O.
  • 5.15: \(\boxed{\ce{Er^{3+}}}\), erbium(III), [Er+3].
  • 5.16: \(\boxed{\lambda=470.0\ \mathrm{nm}}\).
  • 5.17: \(\boxed{\text{Blue-green}}\), taking the emitted light to correspond to the same electronic transition that the garnet absorbs.

Solution

The garnets and metal oxides are extended crystalline solids, so a molecular SMILES would not describe their structures. The atomic-ion SMILES above identify the relevant ions but do not encode their coordination geometry or the shared oxygen lattice. The oxygen sites are formally oxide ions, \(\ce{O^{2-}}\), [O-2].

5.1 — Density of almandine

Using standard atomic masses, \[ \begin{aligned} M\bigl(\ce{Fe3Al2(SiO4)3}\bigr) &=3M(\ce{Fe})+2M(\ce{Al})+3M(\ce{Si})+12M(\ce{O})\\ &=497.74\ \mathrm{g\,mol^{-1}}. \end{aligned} \]

Each unit cell contains eight formula units, so \[ m_{\mathrm{cell}}=\frac{8M}{N_A}. \]

The cell volume is \[ V_{\mathrm{cell}}=a^3 =(11.50\times10^{-8}\ \mathrm{cm})^3 =1.520875\times10^{-21}\ \mathrm{cm^3}. \]

Therefore, \[ \rho= \frac{8(497.74\ \mathrm{g\,mol^{-1}})} {(6.02214076\times10^{23}\ \mathrm{mol^{-1}}) (1.520875\times10^{-21}\ \mathrm{cm^3})} =\boxed{4.348\ \mathrm{g\,cm^{-3}}}. \]

5.2 — Octahedral chromium(III)

Chromium(III) has three \(d\) electrons. In an octahedral field, the three \(t_{2g}\) orbitals lie below the two \(e_g\) orbitals. Hund’s rule places one electron in each lower orbital.

The required drawing has:

Level Orbitals Electron filling
Upper \(e_g\) \(d_{z^2},d_{x^2-y^2}\) Both empty
Lower \(t_{2g}\) \(d_{xy},d_{xz},d_{yz}\) \(\uparrow,\ \uparrow,\ \uparrow\)

Thus, \[ \boxed{t_{2g}^{3}e_g^{0}},\qquad S=\frac32. \]

The six oxygen neighbours lie along the three mutually perpendicular coordination axes. This geometry is not represented by the central-ion SMILES.

5.3 — Octahedral trivalent ion with a diamagnetic low-spin state

The required configuration is \(d^6\):

  • Low-spin: all six electrons occupy \(t_{2g}\), producing three pairs.
  • High-spin: \(t_{2g}^{4}e_g^{2}\), producing four unpaired electrons.

For a first-row trivalent ion, \(d^6\) corresponds to cobalt(III): \[ \boxed{\ce{Co^{3+}}}. \]

A \(d^0\) ion would be diamagnetic in either arrangement and therefore would not satisfy the question.

5.4 — Dodecahedral iron(II)

Iron(II) has six \(d\) electrons.

Because the answer-sheet diagram is missing, describe the usual dodecahedral scheme by energy order: a lowest singlet \(a\), a second singlet \(b\), a degenerate pair \(c,d\), and a highest singlet \(e\).

The required fillings are:

Level, highest first Lower-spin arrangement High-spin arrangement
\(e\) Empty \(\uparrow\)
\(c,d\) \(\uparrow,\ \uparrow\) \(\uparrow,\ \uparrow\)
\(b\) \(\uparrow\downarrow\) \(\uparrow\)
\(a\) \(\uparrow\downarrow\) \(\uparrow\downarrow\)

Consequently, \[ \begin{aligned} \text{Lower-spin: }&a^2b^2c^1d^1,\qquad S=1,\\ \text{High-spin: }&a^2b^1c^1d^1e^1,\qquad S=2. \end{aligned} \]

Both are paramagnetic. Pairing the two electrons within just one of the degenerate \(c,d\) orbitals would add pairing energy without lowering their orbital energy, so it is unfavourable.

5.5 — Pairing-energy inequalities

Define \[ g_L=\varepsilon_b-\varepsilon_a,\qquad g_M=\varepsilon_c-\varepsilon_b,\qquad g_H=\varepsilon_e-\varepsilon_c. \]

Counting one pairing penalty \(P\) for every doubly occupied orbital gives \[ U_{\mathrm{LS}} =2\varepsilon_a+2\varepsilon_b+2\varepsilon_c+2P \] and \[ U_{\mathrm{HS}} =2\varepsilon_a+\varepsilon_b+2\varepsilon_c+\varepsilon_e+P. \]

Their difference is \[ U_{\mathrm{HS}}-U_{\mathrm{LS}} =\varepsilon_e-\varepsilon_b-P =g_M+g_H-P. \]

Thus, \[ \boxed{ \begin{array}{ll} P<g_M+g_H &: \text{lower-spin arrangement},\\ P>g_M+g_H &: \text{high-spin arrangement}. \end{array}} \]

At equality, the two arrangements have equal energy in this model.

The physical threshold is therefore the separation between the highest and second-lowest orbitals. Its expression in the missing diagram’s symbols is:

Gap labelling, bottom to top Lower-spin condition High-spin condition
\(E_1,E_2,E_3\) \(P<E_2+E_3\) \(P>E_2+E_3\)
\(E_3,E_2,E_1\) \(P<E_1+E_2\) \(P>E_1+E_2\)

5.6 — Diamagnetic divalent ions in the dodecahedral field

An odd number of electrons cannot give a diamagnetic ion. For the even electron counts, the possible closed lower levels are:

Ion \(d\) count Diamagnetic filling Condition for that filling
Titanium(II), \(\ce{Ti^{2+}}\) \(d^2\) \(a^2\) \(P<g_L\)
Chromium(II), \(\ce{Cr^{2+}}\) \(d^4\) \(a^2b^2\) \(P<g_M\)
Nickel(II), \(\ce{Ni^{2+}}\) \(d^8\) \(a^2b^2c^2d^2\) \(P<g_H\)

Each has a paramagnetic high-spin arrangement. The \(d^6\) ion remains paramagnetic in its lower-spin arrangement, as established in 5.4; a \(d^{10}\) ion remains diamagnetic.

The stated condition \(P>E_3\) excludes whichever closed-shell arrangement requires \(P<E_3\):

  • If \(E_3=g_H\), the answer is titanium and chromium.
  • If \(E_3=g_L\), the answer is chromium and nickel.
  • If \(E_3=g_M\), the answer is titanium and nickel.

The original diagram is needed to select the intended case without assuming its labels.

5.7 — Assignment of absorption spectra

The observed colour comes from visible wavelengths that are comparatively weakly absorbed.

Spectrum Mineral Reason
A Citrine Strong absorption in the violet–blue region leaves predominantly yellow-orange light.
B Bohemian garnet Strong blue and green absorption, with much weaker red absorption, produces a red appearance.
C Uvarovite Absorption in violet/blue and red regions leaves a transmission window in the green.
D Sapphire Strong absorption around the yellow–orange/red region leaves blue light relatively unabsorbed.

5.8 — Garnet under monochromatic blue-green illumination

Blue-green light falls in a strongly absorbed region of the garnet spectrum. Under the intended ordinary-colour model, little light emerges, and there is no incident red component to provide the usual red appearance.

The answer is therefore \[ \boxed{\text{Black}}. \]

5.9 — Coupled substitution in schorlomite

Replacing iron(III), \(\ce{Fe^{3+}}\), with titanium(IV), \(\ce{Ti^{4+}}\), [Ti+4], increases the positive charge by one per substitution.

Replacing silicon(IV), \(\ce{Si^{4+}}\), [Si+4], with iron(III) decreases it by one. Charge neutrality therefore requires equal numbers of these two substitutions.

There are two octahedral iron sites and three tetrahedral silicon sites per formula unit. Thus, \[ n(\text{octahedral substitutions})=0.05\times2=0.10. \]

The same \(0.10\) tetrahedral sites must be substituted, giving \[ \boxed{\frac{0.10}{3}\times100\%=3.33\%}. \]

The corresponding site composition can be written \[ \ce{Ca3(Fe1.90Ti0.10)(Si2.90Fe0.10)O12}, \] where the first parenthesis describes octahedral sites and the second describes tetrahedral sites.

5.10 — Iron(III) in octahedral and tetrahedral sites

Iron(III) is \(d^5\). The weak oxygen-ligand field, especially in tetrahedral coordination, gives the high-spin arrangement. Both chromophores consequently have five unpaired electrons, consistent with the statement.

For the octahedral chromophore, draw:

Level Filling
Upper \(e_g\), two orbitals \(\uparrow,\ \uparrow\)
Lower \(t_{2g}\), three orbitals \(\uparrow,\ \uparrow,\ \uparrow\)

Hence, \[ \boxed{t_{2g}^{3}e_g^{2}}. \]

For the tetrahedral chromophore, the order reverses: the two \(e\) orbitals are lower, and the three \(t_2\) orbitals are higher.

Level Filling
Upper \(t_2\), three orbitals \(\uparrow,\ \uparrow,\ \uparrow\)
Lower \(e\), two orbitals \(\uparrow,\ \uparrow\)

Hence, \[ \boxed{e^{2}t_2^{3}}. \]

Both have \(S=5/2\). The tetrahedral drawing represents four oxygen neighbours at the vertices of a tetrahedron; the octahedral drawing represents six neighbours along the coordination axes.

5.11 — Pairing energy and crystal-field splittings

In each high-spin \(d^5\) ground state, every orbital contains one electron. The relevant excitation transfers an electron from an upper orbital into a lower orbital, where it must form a pair.

This lowers the one-electron orbital energy by \(\Delta\), but costs pairing energy \(P\). Its net energy is therefore \[ E_{\mathrm{transition}}=P-\Delta. \]

For the octahedral ion, the configuration changes as \[ t_{2g}^{3}e_g^{2}\longrightarrow t_{2g}^{4}e_g^{1}. \]

For the tetrahedral ion, \[ e^{2}t_2^{3}\longrightarrow e^{3}t_2^{2}. \]

Each excited configuration has three unpaired electrons rather than five. The change in total spin explains why these transitions are weak.

Define spectroscopic energies \[ \widetilde P=\frac{P}{hc},\qquad \widetilde\Delta=\frac{\Delta}{hc}. \]

The given transitions require \[ \widetilde P-\widetilde\Delta_{\mathrm{oct}} =11\,000\ \mathrm{cm^{-1}}, \] \[ \widetilde P-\widetilde\Delta_{\mathrm{tet}} =22\,000\ \mathrm{cm^{-1}}, \] with \[ \widetilde\Delta_{\mathrm{tet}} =\frac49\widetilde\Delta_{\mathrm{oct}}. \]

Subtracting the first equation from the second, \[ \left(1-\frac49\right)\widetilde\Delta_{\mathrm{oct}} =11\,000\ \mathrm{cm^{-1}}. \]

Therefore, \[ \boxed{\widetilde\Delta_{\mathrm{oct}} =19\,800\ \mathrm{cm^{-1}}}, \] \[ \boxed{\widetilde\Delta_{\mathrm{tet}} =8\,800\ \mathrm{cm^{-1}}}, \] and \[ \boxed{\widetilde P =11\,000+19\,800 =30\,800\ \mathrm{cm^{-1}}}. \]

Since \(P\) exceeds both splittings, the assumed high-spin ground states are consistent.

5.12 — Cation positions in YAG

Yttrium(III) is substantially larger than aluminium(III). It therefore occupies the larger, eight-coordinate dodecahedral A sites.

Aluminium(III) occupies both the six-coordinate octahedral B sites and the four-coordinate tetrahedral sites normally occupied by silicon.

The site assignment is \[ \boxed{\ce{Y3Al2(AlO4)3}}, \] equivalent to \(\ce{Y3Al5O12}\).

5.13 — Five-percent cerium substitution

Interpreting the missing formula as \(\ce{Y_xCe_yAl5O12}\), there are three yttrium-type sites per formula unit: \[ x+y=3. \]

Replacing 5% of them with cerium(III) gives \[ y=0.05(3)=0.15,\qquad x=3-0.15=2.85. \]

Thus, \[ \boxed{x=2.85,\qquad y=0.15}, \] and the composition is \[ \boxed{\ce{Y2.85Ce0.15Al5O12}}. \]

The substitution is isovalent, so it requires no compensating charge defect.

5.14 — Preparation of cerium-doped YAG

The reactants are yttrium(III) oxide, aluminium oxide, cerium(IV) oxide, and hydrogen. Cerium is reduced from \(+4\) in \(\ce{CeO2}\) to \(+3\) in the garnet, while hydrogen forms water.

For one formula unit, the balance is \[ \ce{1.425Y2O3 + 2.5Al2O3 + 0.15CeO2 + 0.075H2 -> Y2.85Ce0.15Al5O12 + 0.075H2O}. \]

Multiplication by 40 gives \[ \boxed{ \ce{57Y2O3 + 100Al2O3 + 6CeO2 + 3H2 -> 40Y2.85Ce0.15Al5O12 + 3H2O} }. \]

Both sides contain 114 Y, 200 Al, 6 Ce, 483 O, and 6 H atoms. The coefficients have no common divisor. In particular, the yttrium oxide coefficient is \(57/40\) times the product coefficient, so 40 is the smallest permitted positive integer product coefficient.

The molecular species are hydrogen, \(\ce{H2}\), [H][H], and water, \(\ce{H2O}\), O.

5.15 — Blue-emitting cation

Blue light near \(450\)–\(495\ \mathrm{nm}\) corresponds approximately to \[ 20\,200\text{–}22\,200\ \mathrm{cm^{-1}}. \]

The diagram’s \(21\,277\ \mathrm{cm^{-1}}\) transition lies in this interval and belongs to \[ \boxed{\ce{Er^{3+}}}. \]

5.16 — Emission wavelength

Wavelength and wavenumber are reciprocals: \[ \lambda=\frac1{\widetilde\nu}. \]

Thus, \[ \lambda =\frac1{21\,277\ \mathrm{cm^{-1}}} =4.69991\times10^{-5}\ \mathrm{cm} =469.991\ \mathrm{nm}. \]

Therefore, \[ \boxed{\lambda\approx470.0\ \mathrm{nm}}. \]

5.17 — Photoluminescence colour

In the intended interpretation, the emission reverses the electronic transition responsible for absorption.

The garnet appears blood-red under white illumination because it absorbs the complementary blue-green component strongly. Reversing that absorption transition emits photons of the corresponding blue-green energy.

The intended answer is therefore \[ \boxed{\text{Blue-green}}. \]

The solution was checked against the official solution.