Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Parameters for the RM1 Quantum Chemical Calculation of Complexes of the Trications of Thulium, Ytterbium and Lutetium

Abstract

The RM1 quantum chemical model for the calculation of complexes of Tm(III), Yb(III) and Lu(III) is advanced. Subsequently, we tested the models by fully optimizing the geometries of 126 complexes. We then compared the optimized structures with known crystallographic ones from the Cambridge Structural Database. Results indicate that, for thulium complexes, the accuracy in terms of the distances between the lanthanide ion and its directly coordinated atoms is about 2%. Corresponding results for ytterbium and lutetium are both 3%, levels of accuracy useful for the design of lanthanide complexes, targeting their countless applications.

Citation: Filho MAM, Dutra JDL, Rocha GB, Simas AM, Freire RO (2016) Parameters for the RM1 Quantum Chemical Calculation of Complexes of the Trications of Thulium, Ytterbium and Lutetium. PLoS ONE 11(5): e0154500. https://doi.org/10.1371/journal.pone.0154500

Editor: Jason B. Love, University of Edinburgh, UNITED KINGDOM

Received: November 23, 2015; Accepted: April 14, 2016; Published: May 25, 2016

Copyright: © 2016 Filho et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The authors have no support or funding to report.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The computational chemistry of lanthanide complexes at the semiempirical level started in 1994 with the introduction of the Sparkle model[1, 2], which allowed, for the first time, fast quantum chemical geometry optimizations of relatively large complexes. The Sparkle model filled the important gap of lanthanide complexes modelling[3–9] by opening the possibility of having the complexes’ UV-Vis spectra [10] and ligand field parameters[11, 12] predicted.

In 2004, the model was improved with the introduction of Gaussian functions in its core-core repulsion [13] to make it consistent with the AM1 semiempirical model [14]. This improved the Sparkle model quite substantially, and, later, in 2005, the Sparkle model was fully parameterized within AM1 for thulium [15], ytterbium [16], and lutetium [17]. Since different semiempirical models have different characteristics and scopes of applications, it soon became clear that there would be value in parameterizing the Sparkle model for all of the most used and widely distributed semiempirical models available. Thus, the Sparkle model was parameterized for PM3 [18–20]; for PM6 [21], the first semiempirical model parameterized for almost all stable atoms of the periodic table; for PM7 [22], with an emphasis on materials and solid state calculations; and for our RM1 [23].

Overall, results indicate that all Sparkle models are very accurate only when all the directly coordinated atoms to the lanthanides in the complex are either an oxygen or a nitrogen–usually, the most common bonding situation. However, there are many instances in which other atoms coordinate directly with the lanthanide trications, such as carbon, sulfur, chlorine, bromine, and iodine. Thulium, for example, makes a number of complexes directly coordinated to carbon atoms, such as thulium alkylidene complexes [24], which contain a thulium carbon double bond character with the participation of a π-overlap between the carbon and the thulium center for the stabilization of the complex. Likewise, there was recently a study of the interaction of a complex with DNA, which had three chlorine atoms directly coordinated to an ytterbium atom[25]. Another example is a bis(alkyl) complex of lutetium attached to suitable ligands, which, upon activation by an organoborate, initiates the living polymerization of isoprene with high activity [26]. Furthermore, very different bonding situations may also arise, as in the synthesis of a mononuclear lutetium imido complex, which involves intermediate complexes with chlorine atoms and a cyclooctatetraenyl ring, all directly coordinated to lutetium [27]. These are examples of bonding situations, which the Sparkle models cannot properly address. All that is within the realm of the diversity of recent advances in the applications of complexes of thulium as near infrared emitters [28–30], as contrast agents for magnetic resonance imaging [31], and as catalyzers [32]. Also within the realm of recent advances is the utilization of complexes of ytterbium, again as near infrared emitters [33–35], as single-molecule magnets [36, 37], as catalysts [38–40], as selective biomarkers for cancer cell imaging [41], and as DNA binders [25]. Finally, within the scope of the recent advances is the use of complexes of lutetium as gas sensors [42], as catalysts [43, 44], in organic films with switchable electronic and/or interface properties with external electric field [45]; and as near infrared absorbing electrochromes [46]. These complexes, however, tend to have poor thermal stability, poor photostability, and low mechanical strength, usually requiring that they be incorporated, for example, either into sol-gel [30], or into ordered mesoporous materials via a covalently bonded group[28, 29]. Consequently, there is a strong need for theoretical modelling methods capable of addressing these challenges. Recently, we introduced a major upgrade to the Sparkle model, in order to arrive at a model capable of describing any type of bonds between a lanthanide metal and a ligand, within the framework of our semiempirical model RM1 (Recife Model 1) [47], we called the “RM1 model for the lanthanides” [48, 49].

Method

RM1 model for the lanthanides

RM1 is a semiempirical molecular orbital model, with the same algebraic structure of AM1 [14], but reparameterized in 2006 with modern numerical techniques [47]. RM1 is accurate and robust for the types of atoms for which it was originally parameterized: H, C, O, N, P, S, F, Cl, Br, and I. Although this set of atoms looks small, these atoms comprise the vast majority of all atoms present in metal ligands and biomolecules. RM1 was therefore our model of choice for parameterizing the lanthanide trications. In the RM1 model for the lanthanides, we then regard the semiempirical lanthanide atom as an amalgamation of two separate entities: the core and the valence shell. The semiempirical core represents the [Xe]4fn electrons, with n = 12 for Tm, n = 13 for Yb, and n = 14 for Lu. The valence shell is described by three sets of atomic orbitals: 5d, 6s, and 6p, and contains 3 valence electrons. Note that this arrangement is capable of describing only the trications of the lanthanides. Therefore, the present parameterization will be only for complexes of trivalent thulium, ytterbium and lutetium. The RM1 parameters are presented in Table 1.

thumbnail
Download:
Table 1. Parameters for the RM1 model for the trications of Tm, Yb and Lu.

https://doi.org/10.1371/journal.pone.0154500.t001

Results and Discussion

Parameterization

In 2006, when we introduced the Sparkle/PM3 model for thulium [18], we further perfected our parameterization procedure to make sure the model would acquire a more robust attribute. So, following this line, we first collected all complexes of Tm(III), Yb(III), and Lu(III) of high crystallographic quality (R<5%) extant in the 2015 release of the Cambridge Crystallographic Database, CSD[50, 51]. Of course, it would be unfeasible to parameterize the method using all ligands found. Therefore, we used a sampling technique in order to pick, from the universe of complexes, two smaller sets to become the parameterization sets. In order to do that, for each of the lanthanide trications, we first associated, to each of its complexes, a number corresponding to a measure of the difficulty of predicting its geometry in order to guarantee that the sets would be balanced between complexes with easy to predict geometries and complexes with those geometries that are more difficult to predict. We chose this number to be a measure of the distance between the crystallographic geometry obtained from the CSD2015 and a fully optimized Sparkle/AM1 geometry. This number, Ri, is defined for each complex i, in Eq 1, below, as: (1) where j is an index that runs over all different types of bonds, for example, Ln-O, Ln-N, Ln-C, etc; k runs over all bonds of type j; σjdist is the standard deviation of all bonds of type j from the universe of complexes; CSD refers to geometric variables, either distances d, or angles θ, obtained from the Cambridge Crystallographic Database; and Calc refers to geometric variables obtained from Sparkle/AM1 calculations; l runs over all angles in complex i and σangle is the standard deviation of all angles from all complexes in the universe of complexes. For the angles, there was no need to separate them into types, because they all form a homogeneous set. Subsequently, we carried out a divisive hierarchical clustering technique DIANA [52] on the complexes and obtained the stratified sampling in the form of a dendogram. We then applied an optimum allocation to it and arrived to two sub-sets from the universe set of complexes: a smaller set, which we called the small set and a larger one, we called the large set. We did that for each of the three lanthanides considered: thulium, ytterbium and lutetium. For thulium, the universe of complexes contains 19 complexes, the small set contains 5 complexes and the large set contains 10 complexes. The respective numbers for ytterbium are 60, 13 and 31; and, for lutetium, 47, 6, and 14. The universe, small and large sets for each of the lanthanides are described in Tables A-C in S1 File.

By using a combination of non-linear numerical optimization techniques for each of the lanthanides, we then first minimized the sum of all Ri calculated for all complexes of the small set–with the only difference that Calc in Eq 1 now refers to the fully optimized geometry of the intermediary model being considered in the optimization step. After this optimization converged, we then proceeded by minimizing the sum of all Ri calculated for all complexes of the large set. We carried out this second minimization of the larger set in order to improve the accuracy of the model. After this second minimization converged, we then considered the RM1 model for each of the lanthanides as terminated. We then proceeded to compute two accuracy measures for the model for each complex i, UMEi, based on the unsigned mean error defined as: (2) where n is the number of j bonds present in complex i, d are bond distances, CSD is the crystallographic bond distance from CSD and RM1 is the distance for the fully optimized geometry of the complex for the RM1 model for the lanthanide being taken into consideration. The first UME took into consideration only distances between the lanthanide ion and the directly coordinating atoms, and are averaged up for all complexes of the universe of complexes and are called UME(Ln-L)s. These lanthanide–directly coordinated ligand atom distances are the most important ones for the calculation of ligand field parameters. The second set of distances, includes all of these plus all distances in the coordination polyhedron, i.e. distances we call L-L’, where L and L’ are any two directly coordinated atoms, averaged out for all complexes of the universe of complexes, and that are called here simply UMEs. All unsigned mean errors are defined mathematically in the interval from zero to infinity. Therefore, in principle, they should follow a gamma distribution function, something that can be verified by means of the one-sample nonparametric Kolmogorov-Smirnoff test, whose p-value must be above 0.05 for the fit of the corresponding UME data to be acceptable within a 95% confidence interval. If the fit passes the test, the mean is statistically justified as an accuracy measure of the model.

As an example, Figs 1 and 2 show histograms of UME(Tm-L)s and UMEs for the thulium model, superimposed to the corresponding fitted gamma distributions. The p-values are, respectively, 0.736 and 0.945 indicating that the mean UME(Tm-L) of 0.050Å, and mean UME of 0.111Å are good accuracy measures of the RM1 model for thulium. Since the thulium-directly coordinated atom distances mostly lie in the 2.3Å to 2.6Å, this implies that the model is accurate to within 2% for these distances. Corresponding mean UME(Yb-L) and UME(Lu-L) data for the RM1 model for Yb(III) and Lu(III), both equal to 0.08Å, imply that these models are accurate to within 3% for these lanthanides.

thumbnail
Download:
Fig 1. Histogram of the UME(Tm-L)s for all 19 complexes of Tm(III) optimized via the RM1 model being advanced in this article.

The mean and variance are obtained from the fitted gamma distribution, and the p-value of the one-sample nonparametric Kolmogorov-Smirnoff is also shown. This value is above 0.05, and therefore the data can be considered adjusted to the fitted gamma distribution within a 95% confidence interval.

https://doi.org/10.1371/journal.pone.0154500.g001

thumbnail
Download:
Fig 2. Histogram of the UMEs for all 19 complexes of Tm(III) optimized via the RM1 model being advanced in this article.

The mean and variance are obtained from the fitted gamma distribution. The p-value of the one-sample nonparametric Kolmogorov-Smirnoff is also shown. This value is above 0.05, and therefore the data can be considered adjusted to the fitted gamma distribution within a 95% confidence interval.

https://doi.org/10.1371/journal.pone.0154500.g002

Comparison with the previous Sparkle models

As mentioned in the introduction, the RM1 model for Tm(III), Yb(III), and Lu(III) complexes, is being presented in this article to expand the applicability of the quantum chemical semiempirical modelling of lanthanides to complexes with directly coordinated atoms other than oxygen or nitrogen.

Fig 3 shows the UME(Tm-L) for each type of directly coordinated atom L, indicated on the horizontal axis. The blue bars on the left side of the double bars represent the mean errors of the RM1 model, while the brown bars on the right side of the double bars represent the range of the mean errors of the various Sparkle models. The bottom part of the light brown bar indicates the error of the most accurate Sparkle model, and the top part of the light brown bar indicates the error of the least accurate of the Sparkle models for the particular type of bond.

thumbnail
Download:
Fig 3. Comparison between the accuracy measures, UME(Tm-L)s, of the RM1 model for Tm(III), being advanced in this article (blue bars) and the various Sparkle models (two-lightness bars) for the various types of directly coordinated bonds, indicated on the horizontal axis by the atom directly coordinated to the metal ion: either O, N, C, Cl, or I.

L refers to the sum of the UMEs of all these atoms, plus UMEs for Tm-Tm bonds. Polyhedron refers to the sum of all UMEs in L plus the sum of all UMEs between any two atoms of the coordination polyhedron. The top of the light orange bar indicates the UME of the least accurate of the Sparkle models, and the bottom of the top light orange bar indicates the UME of the most accurate of the Sparkle models.

https://doi.org/10.1371/journal.pone.0154500.g003

Figs 4 and 5 show corresponding figures for ytterbium and lutetium, respectively. By examining the three figures, one can immediately note that all Sparkle models are indeed accurate for Ln-O and Ln-N bonds, for all three lanthanides. However, the errors significantly increase when the metals are coordinated to a carbon atom or to a chlorine. The errors then become unacceptably large when the directly coordinated atom to the lanthanide is either S, Br or I.

thumbnail
Download:
Fig 4. Comparison between the accuracy measures, UME(Yb-L)s, of the RM1 model for Yb(III), being advanced in this article (blue bars) and the various Sparkle models (two-lightness bars) for the various types of directly coordinated bonds, indicated on the horizontal axis by the atom directly coordinated to the metal ion: either O, N, C, S, Cl, or Br.

L refers to the sum of the UMEs of all these atoms, plus UMEs for Tm-Tm bonds. Polyhedron refers to the sum of all UMEs in L plus the sum of all UMEs between any two atoms of the coordination polyhedron. The top of the light orange bar indicates the UME of the least accurate of the Sparkle models, and the bottom of the top light orange bar indicates the UME of the most accurate of the Sparkle models.

https://doi.org/10.1371/journal.pone.0154500.g004

thumbnail
Download:
Fig 5. Comparison between the accuracy measures, UME(Lu-L)s, of the RM1 model for Lu(III), being advanced in this article (blue bars) and the various Sparkle models (two-lightness bars) for the various types of directly coordinated bonds, indicated on the horizontal axis by the atom directly coordinated to the metal ion: either O, N, C, S, Cl, or Br.

L refers to the sum of the UMEs of all these atoms, plus UMEs for Tm-Tm bonds. Polyhedron refers to the sum of all UMEs in L plus the sum of all UMEs between any two atoms of the coordination polyhedron. The top of the light orange bar indicates the UME of the least accurate of the Sparkle models, and the bottom of the top light orange bar indicates the UME of the most accurate of the Sparkle models.

https://doi.org/10.1371/journal.pone.0154500.g005

Tables 2–4 show the raw data used to build Figs 3–5. Take the case of Tm-I bonds in Table 2. The most accurate sparkle model is Sparkle/PM6 with an UME(Tm-I) = 0.4345Å, whereas the least accurate is Sparkle/PM7 with a corresponding value of 1.6763Å. This is what is indicated in Fig 3 by the light brown bar over the symbol I.

thumbnail
Download:
Table 2. RM1, Sparkle/AM1, Sparkle/PM3, Sparkle/PM6, Sparkle/PM7 and Sparkle/RM1 Unsigned Mean Errors (Å) for Thulium.

https://doi.org/10.1371/journal.pone.0154500.t002

thumbnail
Download:
Table 3. RM1, Sparkle/AM1, Sparkle/PM3, Sparkle/PM6, Sparkle/PM7 and Sparkle/RM1 Unsigned Mean Errors for Ytterbium.

https://doi.org/10.1371/journal.pone.0154500.t003

thumbnail
Download:
Table 4. RM1, Sparkle/AM1, Sparkle/PM3, Sparkle/PM6, Sparkle/PM7 and Sparkle/RM1 Unsigned Mean Errors for Lutetium.

https://doi.org/10.1371/journal.pone.0154500.t004

Furthermore, on Table 3, one can see that Yb-C bonds are very common. Actually, bonds of the type Yb-C are more numerous (242 bonds) in the universe set of Yb complexes in CSD than bonds of the types Yb-O (231 bonds) or Yb-N (97 bonds). The same trend occurs for lutetium as one can clearly see from Table 4. The larger errors of the previous sparkle models occur for sulfur, bromine and iodine. However, complexes with these atoms directly coordinated to the lanthanides are rare. So much so, that we could not even find any such case in the universe set of complexes of lutetium.

The RM1 model calculates isolated structures. However, due to its semiempirical character, it is nevertheless able to predict crystallographic structures with high accuracy, implicitly taking into account solid state effects, such as packing effects.

Conclusion

Results indicate that the present RM1 models for thulium, ytterbium, and lutetium do indeed correct inadequacies of the previous Sparkle models, especially for ytterbium and lutetium, where mainly bonds with carbon atoms directly coordinated to the lanthanide ion are very common.

In conclusion, if the complex of interest has any directly coordinated atoms other than oxygen or nitrogen, then the usage of the present RM1 model for thulium, ytterbium and lutetium is indispensable.

Supporting Information

S1 File. Instructions to run the RM1 calculations.

Instructions on how to run the RM1 model for the lanthanides in MOPAC2012, together with sample calculations on complexes of each of the parameterized lanthanide trications: Tm(III), Yb(III), and Lu(III).

https://doi.org/10.1371/journal.pone.0154500.s001

(DOCX)

Author Contributions

Wrote the paper: AMS ROF. Conceived the model: ROF GBR AMS. Coded the model into MOPAC: GBR. Conceived the parameterization techniques: AMS GBR. Conceived the sampling of the reference structures and the statistical validation of the model: AMS. Prepared the data and carried out the parameterization: JDLD MAMF ROF.

References

  1. 1. de Andrade AVM, da Costa NB, Simas AM, de Sa GF. Sparkle Model for the Quantum-Chemical Am1 Calculation of Europium Complexes. Chem Phys Lett. 1994;227(3):349–53. pmid:ISI:A1994PF45600019.
  2. 2. de Andrade AVM, da Costa NB, Simas AM, de Sa GF. Sparkle Model for the Quantum-Chemical Am1 Calculation of Europium Complexes of Coordination-Number-9. J Alloy Compd. 1995;225(1–2):55–9. pmid:ISI:A1995RR58500013.
  3. 3. de Sa GF, Malta OL, Donega CD, Simas AM, Longo RL, Santa-Cruz PA, et al. Spectroscopic properties and design of highly luminescent lanthanide coordination complexes. Coordin Chem Rev. 2000;196:165–95. pmid:ISI:000085008400007.
  4. 4. da Costa NB, Freire RO, dos Santos MAC, Mesquita ME. Sparkle model and intensity parameters of the Eu(3-amino-2-carboxypyridine-N-oxide)(3)3H2O complex. J Mol Struc-Theochem. 2001;545:131–5. pmid:ISI:000170092600013.
  5. 5. de Mesquita ME, Silva FRGE, Albuquerque RQ, Freire RO, da Conceicao EC, da Silva JEC, et al. Eu(III) and Gd(III) complexes with pirazyne-2-carboxylic acid: luminescence and modelling of the structure and energy transfer process. J Alloy Compd. 2004;366(1–2):124–31. pmid:ISI:000189081000024.
  6. 6. Freire RO, Silva FRGE, Rodrigues MO, de Mesquita ME, Junior NBD. Design of europium(III) complexes with high quantum yield. J Mol Model. 2005;12(1):16–23. pmid:ISI:000233482800003.
  7. 7. Freire RO, Albuquerque RQ, Junior SA, Rocha GB, de Mesquita ME. On the use of combinatory chemistry to the design of new luminescent Eu3+ complexes. Chem Phys Lett. 2005;405(1–3):123–6. pmid:ISI:000228078700023.
  8. 8. de Mesquita ME, Junior SA, Oliveira FC, Freire RO, Junior NBC, de Sa GF. Synthesis, spectroscopic studies and structure prediction of the new Tb(3-NH2PIC)(3)center dot 3H(2)O complex. Inorg Chem Commun. 2002;5(4):292–5. doi: Pii S1387-7003(02)00382-9 pmid:ISI:000175179000016.
  9. 9. Borges AS, Dutra JDL, Freire RO, Moura RT, Da Silva JG, Malta OL, et al. Synthesis and Characterization of the Europium(III) Pentakis(picrate) Complexes with Imidazolium Countercations: Structural and Photoluminescence Study. Inorg Chem. 2012;51(23):12867–78. pmid:ISI:000311772600031.
  10. 10. de Andrade AVM, Longo RL, Simas AM, de Sa GF. Theoretical model for the prediction of electronic spectra of lanthanide complexes. J Chem Soc Faraday T. 1996;92(11):1835–9. pmid:ISI:A1996UR34900003.
  11. 11. Faustino WM, Rocha GB, Silva FRGE, Malta OL, de Sa GF, Simas AM. Design of ligands to obtain lanthanide ion complexes displaying high quantum efficiencies of luminescence using the sparkle model. J Mol Struc-Theochem. 2000;527:245–51. pmid:ISI:000089238700026.
  12. 12. Freire RO, Rocha GB, Albuquerque RQ, Simas AM. Efficacy of the semiempirical sparkle model as compared to ECP ab-initio calculations for the prediction of ligand field parameters of europium(III) complexes. J Lumin. 2005;111(1–2):81–7. pmid:ISI:000226153300010.
  13. 13. Rocha GB, Freire RO, da Costa NB, de Sá GF, Simas AM. Sparkle Model for AM1 Calculation of Lanthanide Complexes:  Improved Parameters for Europium. Inorg Chem. 2004;43(7):2346–54. pmid:15046511
  14. 14. Dewar MJS, Zoebisch EG, Healy EF, Stewart JJP. The Development and Use of Quantum-Mechanical Molecular-models. 76. AM1—A New General-Purpose Quantum-Mechanical Molecular-Model. J Am Chem Soc. 1985;107(13):3902–9. pmid:WOS:A1985ALC3500024.
  15. 15. Freire RO, Rocha GB, Simas AM. Modeling rare earth complexes: Sparkle/AM1 parameters for thulium (III). Chem Phys Lett. 2005;411(1–3):61–5. pmid:ISI:000230980700012.
  16. 16. Freire RO, Rocha GB, Simas AM. Modeling lanthanide complexes: Sparkle/AM1 parameters for ytterbium (III). J Comput Chem. 2005;26(14):1524–8. pmid:ISI:000231887600010.
  17. 17. Freire RO, da Costa NB, Rocha GB, Simas AM. Sparkle/AM1 structure modeling of lanthanum (III) and lutetium (III) complexes. J Phys Chem A. 2006;110(17):5897–900. pmid:ISI:000237442600039.
  18. 18. Freire RO, Rocha GB, Simas AM. Modeling rare earth complexes: Sparkle/PM3 parameters for thulium(III). Chem Phys Lett. 2006;425(1–3):138–41. pmid:ISI:000238727200029.
  19. 19. Freire RO, Rocha GB, Simas AM. Sparkle/PM3 parameters for praseodymium(III) and ytterbium(III). Chem Phys Lett. 2007;441(4–6):354–7. pmid:ISI:000247985400036.
  20. 20. da Costa NB, Freire RO, Simas AM, Rocha GB. Structure modeling of trivalent lanthanum and lutetium complexes: Sparkle/PM3. J Phys Chem A. 2007;111(23):5015–8. pmid:ISI:000247034000017.
  21. 21. Freire RO, Simas AM. Sparkle/PM6 Parameters for all Lanthanide Trications from La(III) to Lu(III). J Chem Theory Comput. 2010;6(7):2019–23. pmid:ISI:000279751500010.
  22. 22. Dutra JDL, Filho MAM, Rocha GB, Freire RO, Simas AM, Stewart JJP. Sparkle/PM7 Lanthanide Parameters for the Modeling of Complexes and Materials. J Chem Theory Comput. 2013;9(8):3333–41. pmid:24163641
  23. 23. Filho MAM, Dutra JDL, Rocha GB, Freire RO, Simas AM. Sparkle/RM1 parameters for the semiempirical quantum chemical calculation of lanthanide complexes. Rsc Adv. 2013;3(37):16747–55.
  24. 24. Cantat T, Jaroschik F, Ricard L, Le Floch P, Nief F, Mezailles N. Thulium alkylidene complexes: Synthesis, X-ray structures, and reactivity. Organometallics. 2006;25(5):1329–32. pmid:WOS:000235691500035.
  25. 25. Moodi A, Khorasani-Motlagh M, Noroozifar M, Niroomand S. Binding analysis of ytterbium(III) complex containing 1,10-phenanthroline with DNA and its antimicrobial activity. J Biomol Struct Dyn. 2013;31(8):937–50. pmid:WOS:000321732100011.
  26. 26. Yao CG, Liu DT, Li P, Wu CJ, Li SH, Liu B, et al. Highly 3,4-Selective Living Polymerization of Isoprene and Copolymerization with epsilon-Caprolactone by an Amidino N-Heterocyclic Carbene Ligated Lutetium Bis(alkyl) Complex. Organometallics. 2014;33(3):684–91. pmid:WOS:000331341900014.
  27. 27. Panda TK, Randoll S, Hrib CG, Jones PG, Bannenberg T, Tamm M. Syntheses and structures of mononuclear lutetium imido complexes with very short Lu-N bonds. Chem Commun. 2007;(47):5007–9. pmid:WOS:000251336600007.
  28. 28. Feng J, Song SY, Fan WQ, Sun LN, Guo XM, Peng CY, et al. Near-infrared luminescent mesoporous MCM-41 materials covalently bonded with ternary thulium complexes. Micropor Mesopor Mat. 2009;117(1–2):278–84. pmid:WOS:000262879700036.
  29. 29. Feng J, Song SY, Xing Y, Zhang HJ, Li ZF, Sun LN, et al. Synthesis, characterization, and near-infrared luminescent properties of the ternary thulium complex covalently bonded to mesoporous MCM-41. J Solid State Chem. 2009;182(3):435–41. pmid:WOS:000264302000004.
  30. 30. Dang S, Sun LN, Zhang HJ, Guo XM, Li ZF, Feng J, et al. Near-infrared luminescence from sol-gel materials doped with holmium(III) and thulium(III) complexes. J Phys Chem C. 2008;112(34):13240–7. pmid:WOS:000258633600030.
  31. 31. Burdinski D, Lub J, Pikkemaat JA, Langereis S, Grull H, ten Hoeve W. The thulium complex of 1,4,7,10-tetrakisi{[N-(1H-imidazol-2-yl)-carbamoyl]methyl}-1,4,7,10-tetraazacyclododecane (dotami) as a ParaCEST contrast agent. Chem Biodivers. 2008;5(8):1505–12. pmid:WOS:000258973500007.
  32. 32. Lukesova L, Ward BD, Bellemin-Laponnaz S, Wadepohl H, Gade LH. High tacticity control in organolanthanide polymerization catalysis: formation of isotactic poly(alpha-alkenes) with a chiral C-3-symmetric thulium complex. Dalton T. 2007;(9):920–2. pmid:WOS:000244253800002.
  33. 33. Li WZ, Li JY, Li HF, Yan PF, Hou GF, Li GM. NIR luminescence of 2-(2,2,2-trifluoroethyl)-1-indone (TFI) neodymium and ytterbium complexes. J Lumin. 2014;146:205–10. pmid:WOS:000330089600034.
  34. 34. Kalinovskaya IV. Luminescence of ytterbium(III) in mixed-ligand compounds with cinnamic acid and neutral phosphorus-containing ligands. Russ J Phys Chem a+. 2014;88(9):1613–5. pmid:WOS:000340367000030.
  35. 35. Pushkarev AP, Ilichev VA, Balashova TV, Vorozhtsov DL, Burin ME, Kuzyaev DM, et al. Lanthanide complexes with substituted naphtholate ligands: extraordinary bright near-infrared luminescence of ytterbium. Russ Chem B+. 2013;62(2):392–7. pmid:WOS:000328852300014.
  36. 36. Jung J, da Cunha TT, Le Guennic B, Pointillart F, Pereira CLM, Luzon J, et al. Magnetic Studies of Redox-Active Tetrathiafulvalene-Based Complexes: Dysprosium vs. Ytterbium Analogues. Eur J Inorg Chem. 2014;(24):3888–94. pmid:WOS:000340940800014.
  37. 37. Pointillart F, Le Guennic B, Cauchy T, Golhen S, Cador O, Maury O, et al. A Series of Tetrathiafulvalene-Based Lanthanide Complexes Displaying Either Single Molecule Magnet or Luminescence-Direct Magnetic and Photo-Physical Correlations in the Ytterbium Analogue. Inorg Chem. 2013;52(10):5978–90. pmid:WOS:000319720400047.
  38. 38. Basalov IV, Rosca SC, Lyubov DM, Selikhov AN, Fukin GK, Sarazin Y, et al. Divalent Heteroleptic Ytterbium Complexes—Effective Catalysts for Intermolecular Styrene Hydrophosphination and Hydroamination. Inorg Chem. 2014;53(3):1654–61. pmid:WOS:000330812700047.
  39. 39. Yang S, Nie K, Zhang Y, Xue MQ, Yao YM, Shen Q. New [ONOO]-Type Amine Bis(phenolate) Ytterbium(II) and -(III) Complexes: Synthesis, Structure, and Catalysis for Highly Heteroselective Polymerization of rac-Lactide. Inorg Chem. 2014;53(1):105–15. pmid:WOS:000329529800020.
  40. 40. Wang JF, Xu F, Cai T, Shen Q. Addition of amines to nitriles catalyzed by ytterbium amides: An efficient one-step synthesis of monosubstituted N-arylamidines. Org Lett. 2008;10(3):445–8. pmid:WOS:000252648200022.
  41. 41. Zhang T, Chan CF, Lan RF, Li HG, Mak NK, Wong WK, et al. Porphyrin-based ytterbium complexes targeting anionic phospholipid membranes as selective biomarkers for cancer cell imaging. Chem Commun. 2013;49(65):7252–4. pmid:WOS:000321956400022.
  42. 42. Ceyhan T, Altindal A, Ozkaya AR, Erbil MK, Bekaroglu O. Synthesis, characterization, and electrochemical, electrical and gas sensing properties of a novel tert-butylcalix[4]arene bridged bis double-decker lutetium(III) phthalocyanine. Polyhedron. 2007;26(1):73–84. pmid:WOS:000243622000012.
  43. 43. Otero A, Lara-Sanchez A, Najera C, Fernandez-Baeza J, Marquez-Segovia I, Castro-Osma JA, et al. New Highly Active Heteroscorpionate-Containing Lutetium Catalysts for the Hydroamination of Aminoalkenes: Isolation and Structural Characterization of a Dipyrrolidinide-Lutetium Complex. Organometallics. 2012;31(6):2244–55. pmid:WOS:000301895900014.
  44. 44. Meyer N, Zulys A, Roesky PW. A chiral-bridged aminotroponiminate complex of lutetium as catalyst for the asymmetric hydroamination. Organometallics. 2006;25(17):4179–82. pmid:WOS:000239536800023.
  45. 45. Lei SB, Deng K, Yang YL, Zeng QD, Wang C, Jiang JZ. Electric driven molecular switching of asymmetric tris(phthalocyaninato) lutetium triple-decker complex at the liquid/solid interface. Nano Lett. 2008;8(7):1836–43. pmid:WOS:000257504500010.
  46. 46. Pushkarev VE, Kalashnikov VV, Trashin SA, Borisova NE, Tomilova LG, Zefirov NS. Bis(tetrabenzotriazaporphyrinato) and (tetrabenzotriazaporphyrinato)(phthalocyaninato) lutetium(III) complexes—novel sandwich-type tetrapyrrolic ligand based NIR absorbing electrochromes. Dalton T. 2013;42(34):12083–6. pmid:WOS:000322780000005.
  47. 47. Rocha GB, Freire RO, Simas AM, Stewart JJP. RM1: A reparameterization of AM1 for H, C, N, O, P, S, F, Cl, Br, and I. J Comput Chem. 2006;27(10):1101–11. pmid:ISI:000238158900007.
  48. 48. Filho MAM, Dutra JDL, Rocha GB, Simas AM, Freire RO. Semiempirical Quantum Chemistry Model for the Lanthanides: RM1 (Recife Model 1) Parameters for Dysprosium, Holmium and Erbium. PLoS ONE. 2014;9(1):e86376. pmid:24497945
  49. 49. Filho MA, Dutra JDL, Cavalcanti HL, Rocha GB, Simas AM, Freire RO. RM1 Model for the Prediction of Geometries of Complexes of the Trications of Eu, Gd, and Tb. J Chem Theory Comput. 2014;10(8):3031–7. pmid:26588274
  50. 50. Allen FH. The Cambridge Structural Database: a quarter of a million crystal structures and rising. Acta Crystallogr B. 2002;58:380–8. pmid:WOS:000176270700010.
  51. 51. Bruno IJ, Cole JC, Edgington PR, Kessler M, Macrae CF, McCabe P, et al. New software for searching the Cambridge Structural Database and visualizing crystal structures. Acta Crystallogr B. 2002;58:389–97. pmid:WOS:000176270700011.
  52. 52. Kaufman L, Rousseeuw PJ. Finding groups in data: an introduction to cluster analysis: John Wiley & Sons; 2009.