Applications Of Single Crystal X Ray Crystallography

Published Date: 02 Nov 2017

Loughborough University

Tetrahalotin complexes: Structure refinement, database searches, and publication in the journal Acta Crystallographica Vol C.

Applications of single crystal X-ray crystallography.

Danielle Pearce

Supervisor: Dr M. R. J. Elsegood

1/1/2013

Table of Contents

Literature review……………………………………………….2

Abstract………………………………………………………...5

Introduction……………………………………………………6

Experimental…………………………………………………...6

Results………………………………………………………….8

Discussion……………………………………………………...10

Conclusion……………………………………………………..11

References……………………………………………………..12

Acknowledgements……………………………………………13

Appendix………………………………………………………13

Literature Review

Identification the internal structure of a substance is necessary for many applications, not least of which to know exactly what that substance is made of and how the atoms are arranged and bound within the substance. One of the most accurate and widely used methods for doing this is single crystal X-ray diffraction and it is the primary method for identifying new materials as bond lengths can be measured to within thousandths of an angstrom and angles to within tenths of a degree. X-ray crystallography determines the positions of atoms and/or molecules within a crystal. By definition a crystal is a substance constituting of a regular, repeating pattern of atoms, ions or molecules[1]; a wide variety of substances form crystals thus extending the application of this method. SCXD utilises measurement of the angles and intensities of an X-ray beam scattered by the atoms within a rotating crystal to produce a three dimensional electron density map corresponding to a reflection from a plane within the crystal[2].

Bruker Corporation is a leading manufacturer of scientific instruments including X-ray diffractometers. The instruments’ have four key components which work together; source, optics, goniometer and detector. The crystallographic analysis depends on X-rays so the source needs to be reliable. To achieve a monochromatic beam the source is combined with X-ray optics which focuses the beam to increase intensity. A range of monochromators are available such as a flat graphite monochromator. The goniometer head which holds he crystal needs to be finely focusable to ensure the crystal is directly in the path of the X-ray beam. Many detectors are based on charge-coupled device (CCD) and the APEX II detector manufactured by Bruker has 15 times the sensitivity of standard CCD detectors[3]. The collected data is used to determine and refine the structure by using the free software SHELX which was developed in the 1960’s by G. M. Sheldrick[4][5][6].. This can also be used to produce schematics and packing plots of the structure.

As with any method of analysis, SCXD has both strengths and limitations. Key strengths include; the lack of need for separate standards; the method is non-destructive; and a detailed crystal structure is obtained inclusive of bond angles and lengths [7]. Key limitations include; need for a single, stable, optically clear sample of sufficient size; long time period of data collection (2-24 hours); and the larger and more complex the unit cell, the poorer the resolution[8][9].

X-ray diffraction has been used to identify the structures of many compounds and complexes including metal halides. Tin halides have a wide range of properties and uses and their syntheses are relatively straightforward. During the reaction of tin with a halogen to form tetrahalotin, the metal tin goes from a zero oxidation state to a plus four oxidation state. Tin halides have been used in a wide range of applications from use as polyethylene- and polystyrene-bound catalysis for aryl and alkyl bromides[10]. Stannic fluoride is a white solid with a melting point of over 700 oC and is commonly used as a decay preventative in toothpaste[11][12]. On contact with air the colourless liquid stannic chloride forms a smoke which is irritating but not deadly and has a stinging odour; this was used during the first word war as a chemical weapon[13][14]. It is also used as a catalyst in Freidel-Crafts reaction for homogeneous cyclisation, alkylation and acylation. SnCl4 is also used in the production of glass bottles and part of a surface deposition mechanism to improve the strength and resistance to abrasion[11]. Stannic iodide is a bright orange solid that forms a rare complex in aqueous solution of hydrochloric acid - the metal hexaiodide[15]. Stannic bromide is a colourless solid with a relatively low melting point. All four stannic halides are readily hydrolysed and are soluble in a wide range of organic solvents. The synthesis of SnCl4, SnBr4 or SnI4 is achieved by direct reaction of the elements; SnCl4 via reaction of chlorosulfonic acid with tin metal and both SnBr4 and SnI4 by reflux of the elements. SnCl4 can then be reacted with anhydrous HF to yield SnF4[11].

Property

SnF4

SnCl4

SnBr4

SnI4

Colour

white

colourless

colourless

orange-brown

Melting point / oC

-

-33.3

31

144

Boiling point / oC

705 (sublimes)

144

205

348

Density / g cm-3

4.78

2.23

3.34

4.56

Sn-X / pm

188 eq, 202 ax

231

244

264

Van der Waal’s radius of X / pm

147

175

183

198

Electronegativity of X

4.0

3.0

2.8

2.7

Table 1. Properties of the tetrahalotin molecules.[11]

From the data in Table 1., it can be seen that there are trends across the tin halide series; with increasing atomic van der Waal’s radius the Sn-X bond lengths become longer. This trend could be due to an increase in steric strain as going from fluorine through to iodine the size of the Van der Waal’s radius of the halogen increases from though to iodine and/or a decrease in electronegativity. A combination of these two affects the bond lengths. For the chlorine, bromine and iodine compounds the density, melting and boiling point temperatures increases with molecular weight, however the fluorine compound bucks this trend by having higher values than all of the others and much higher than that predicted by the trend. This is due to the nature of the geometric structure of the substance; the tin fluoride is constituted of planar layers of tin atoms with octahedral coordination. These vertex-sharing octahedra contain equatorial fluorine and the remaining two unshared terminal fluorine atoms are trans to each other, whereas the other halides are tetrahedrally coordinated[16]. The two different tin-fluorine bond lengths reflect the two different fluorine environments. In this series the crystal structure determination has been used to identify the structure of the compounds which has enabled understanding of why fluorine doesn’t follow the trends predicted. This is just one of many applications.

Another method of structural study of tin is 119Sn Mossbauer spectroscopy which takes into account parameters including quadrupole splitting (QS) Δ and isomer shift (IS) δ. The QS reflects the field gradient around the tin nucleus due to the π and σ bonds made by the six coordinating atom. This provides information about the Sn-L or Sn-X bond length in regard to t he octahedral compound. The IS parameter defines the measure of s electron density at the tin nucleus and is affected by the electronegativities of any bound atoms as it increases as the covalent character of the Sn-X bond increases (see table 2)[32]. Fluorine has a higher electronegativity than chlorine so the Sn-F bond is less covalent than the Sn-Cl bond resulting in a lower IS. This trend continues throughout the halide series. As the coordination number of tin increases from 4 to 6 the IS value decreases. This can be seen when SnBr4 goes from a coordination number of four to SnBr4(DMF)2 with a coordination number of 6 the IS falls from 1.15 to 0.66; again due to a decrease in covalent character.

Compound

QS / mm s-1

IS / mm s-1

SnF4

1.66

-0.47

SnCl4

0

0.85

SnBr4

0

1.15

SnI4

0

1.55

trans-SnCl4(dmso)2

0.57

0.41

cis-SnCl4(dmso)2

0.41

0.40

trans-SnBr4(DMF)2

0.83

0.66

cis-SnBr4(DMF)2

0.44

0.66

Table 2. Quadrupole splitting and isomer shift values for a range of compounds[32].

Further reactions of the tetrahalotin with bidentate ligands, such as 2,2-bipyridine and 1,10-phenanthroline, take the tin from the +4 oxidation state to the coordinatively saturated +6 state with octahedral geometry. These adduct reactions can occur because tin has vacant d-orbitals which lay low enough in energy to accept a lone pair of electrons form the ligand[11]. Many of these ligands have hard donor atoms such as oxygen or nitrogen but examples with soft donors such as phosphorous have also been synthesised. Usually SnX4 behaves as a Lewis acid, however due to the different structure of SnF4 it is much less reactive and insoluble in weak donor solvents compared to the other three halides which are soluble even in non-coordinating solvents[17].

The structures and functionalities of 2,2-bipyridine and 1,10-phenanthroline are similar and both can undergo a 1:1 reaction with SnX4 to produce a bidentate complex, the former with trans geometry and the latter with cis geometry[11]. [SnF4(2,2-bipyridine)] is remarkably insensitive to water and is insoluble in non-polar to moderately polar solvents. The structure is octahedral and monomeric and has no intermolecular C-H…I hydrogen bonds[18].

[SnI4(2,2-bipyridine)] does not dissolve in standard solvents[19]. The packing plot of [SnI4(1,10-phenanthroline)] shows weak C-H…I hydrogen bonding between the iodine atoms in one molecule and hydrogen atoms in the 1,10-phenanthroline of another molecule. These weak C-H…I interactions hold together the three dimensional network of molecules. Intermolecular H…I bonds are between 3.15 and 3.34 Å in length but it has been shown that [SnI4(1,10-phenanthroline)] contains intramolecular H…I bonds of between 2.95 and 3.00 Å. The van der Waal's radius of iodine is 1.98Å and the packing plot gives the I…I distances as approximately 4Å which just outside the range where I…I iodine interactions would be expected[19]. The van der Waal's radii of fluorine, chlorine and bromine are 1.50Å, 1.80Å, and 1.90Å respectively. Due to the off-set nature of the packing the molecules are not stacked so there are no π-π stacking interactions.

The Cambridge Structure Database (CSD) contains structural database was established in 1965 and is the world repository for crystal structures. It contains data from over five hundred thousand X-ray and neutron diffraction analysis and is used frequently by a range of scientist. Each entry has full details of the paper in which the structure was published along with an interactive 3D model from which bond angles and lengths can be measured; and the structures crystallographic data including space group, unit cell dimensions etc.

X. Wang [21] investigated the reaction of stannic iodide with a number of ligands including the undergraduate part A adduction of triphenylphosphine. On reaction the product is cis-[SnI4(Ph3PO)2] not the [SnI4(Ph3P)2] as expected however it can be obtained if reaction takes place under a nitrogen atmosphere but it is highly air and moisture sensitive. This shows that the tin has a greater affinity for oxygen than it does for phosphorous and will grab any available oxygen to form the phosphine oxide complex. Another unexpected product from this study is that of the stannic iodide and pyridine. Pyridine is an monodentate ligand and on reaction with stannic iodide it is expected that the [SnI4(pyridine)2] complex forms, however this is not the case and instead dipyridinuim iodide triiodide [(pyH)2+I-I3-] is formed instead. The hydrogen atoms in the pyridine ring form both C-H…I and N-H…I hydrogen bonds with both iodine species holding the structure together in zigzagging chains which are cross linked via I-I-I…I weak interactions. This is unexpected at the tin is no longer part of the complex. This study then looks at bidentate ligands that coordinate through both nitrogen and oxygen as these have also been studies previously and many can be found on the CSD[20]. It was found that some of these ligands such as 2, 3-pyazine carboxylic acid do not dissolve in standard solvents so the reactions

Abstract

The aim of this project was to re-refine and a tin and a polyiodide crystal structure to 2013 standards; to search the CSD and then the chemical and crystallographic literature for related examples and to comparisons; and to prepare draft manuscripts and associated figures as schemes for publication in Act Crystallographica, Vol C. All of these aims have been met in this project and the paper. The structure has been refined and the manuscripts have been submitted for publication. During the reaction of tetrahalotin with 2,2-bipyridine and 1,10-phenanthroline take the tin from the plus four oxidation state to the plus six oxidation state. This project reports on findings correlated from previous studies and shows trends within the tetrahalotin complex series coordinated to 1,10-phananthroline and 2,2’-bipyridine. It can be determined that the bond lengths and angles between tin, the halogen and the coordinating atoms are effected by the both the size of the halogen and by the electronegativities of the atoms involved. This study also highlights the difficulties in growing crystals suitable for single crystal X-ray crystallography.

Introduction

As with most modern chemistry, the work undertaken in this project was a continuation and development of previous work. Previous work carried out by a former Loughborough University student follows on from the part A SnI4 and triphenylphosphine reaction and looks at a range of tetraiodotin complexes; their synthesis and properties, but as individual complexes[19]. The bond lengths and angles between tin, the halogen and the coordinating atoms are affected by the both the size of the halogen and by the electronegativities of the atoms involved[11][22]. These values can be obtained accurately using single crystal X-ray diffraction and this study highlights some difficulties in obtaining suitable crystals[23]. The fluorine and chlorine analogues have also been studied previously and many have been published in Acta Crystallographica[19][24]. This project looks at the bromine analogues and comparisons between the whole tetrahalotin series. Tin tetrahalide is known to form bidentate complexes with a variety of monodentate and bidentate ligands in the ratios 1:2 and 1:1 respectively. This provides ample opportunity to study how varying the halogen affects the complex.

Experimental

As a starting point to the project and an example of how the process works crystallographic data from an unrelated sample was collected using the Bruker single crystal X-ray diffractometer under the supervision of Dr Elsegood. To obtain diffraction data the X-ray generator and is set at 50kV and 40 mA. A chiller uses water to cool the generator and X-ray tube. A graphite crystal produces a 0.5 mm beam of X-ray radiation which then masses through a collimator to the crystal. A suitable crystal needs to be roughly 0.5 mm in all dimensions and is then mounted on to the goniometer head using perfluoropolyether oil as a glue. This glue works because it is viscous enough to hold the crystals firmly in place but it is inert and does not itself diffract X-rays, so will not interfere with the data obtained. The crystal is then cooled to 150 K in order to minimise the atoms vibrating which can affect the diffraction pattern obtained. The crystal is positioned in the centre of the X-ray beam by adjusting the goniometer head. The program Apex2 server is used for an initial run to check if it will diffract. An initial 10 second exposure is taken 0.5 degrees through omega (2 x 5 seconds) to give an image. The diffracted radiation is then picked up on a CCD (charged coupled device) detector which is cooled to -50 oC to prevent electrons in the detector jumping between wells and blurring the image. This should preferably give bright round peaks (spots) up to 50 Bragg angle; the number of which depends on the size of the unit cell (larger gives more peaks) and the symmetry (lower symmetry gives fewer peaks) amongst other things. Needle crystals may stick or form together resulting in an effect called twinning. Some may also be curved or form thin sheets, however big block shaped crystals often give the best diffraction patterns. If a choice of crystals is available then it may be beneficial to try the initial run with a range of crystals to compare the initial data as one may produce a more suitable/usable diffraction pattern. Once a suitable crystal is chosen the data collection is run over a number of hours. The structure is then determined and refined using the SHELEX software. To bring the crystal structure of the previously refined tin tetraiodide up to modern standards, a refinement was carried out using SHELX refinement software, data such as bond angles and lengths recorded and figures produced [6]. This refined crystal data was then transferred to a .cif file which will form the manuscript for publishing.

As the main aim of this project is to produce a paper to be published, that brings together and advances on information already obtained about tetrahalotin complexes. A Conquest database a search was carried out to identify possibly tetrahalotin complex series that could be compared.

Many fluoro-, chloro- and iodo- complexes had previously been studies so suitable comparative structures where identified and their methods of synthesis recorded in order to try to replicate these methods with the bromine analogue. To start tin(IV) bromide was synthesised from the pure elements and reacted with a selection of other reagents to form a range of complexes for comparison, as detailed below.

Tetrabromotin Preparation

Acetic acid (25ml) and acetic anhydride (25ml) were placed in an oven-dried 100ml round-bottomed flask. Tin sheet (0.501g), cut into small pieces, and bromine (1.346g) were added to the flask. A reflux condenser with a drying tube was connected and then the mixture heated gently until a vigorous reaction began. This was left to boil until all, or most, of the tin metal had been consumed in the reaction (2 hours and 56 minutes). The solvent was evaporated off until only a very small amount remained and then the white powder was filtered under suction. This crude product was recrystallized from chloroform to give the pure product which was again filtered by suction and then air-dried (30 minutes).

The first attempt at this synthesis was unsuccessful, however two subsequent attempts were successful and more SnBr4 was made as required.

For another comparison further complexes were investigated. As both the SnX4(1,10-phenanthroline)and SnX4(2,2’-bipyridine) complexes have already been synthesised where X is fluorine[18] [19], chlorine[24][25] and iodine[26] [21] and the bromine analogue of SnX4(2,2’-bipyridine) (as found in the Conquest database), the bromine analogue of the SnX4(1,10-phenanthroline) was produced and a comparison was made of the properties and key bond angles and lengths. This information was obtained from the published papers as detailed on Conquest.

Reaction of tetrabromotin with 1,10-phenanthroline

Tetrabromotin (0.438g) and 1,10-phenanthroline (0.198g) where dissolved separately in chloroform (10ml and 6.5ml respectively). The two solutions were mixed in a 100ml round bottomed flask fitted with a reflux condenser. The mixture of solutions was stirred (15 minutes) and after allowing to cool the pale peach-coloured powder precipitate was filtered and dried under vacuum (10 minutes). Attempts at recrystallization were made via both recrystallizations from hot methanol in an ice bath and slow evaporation over one week. These were unsuccessful as the crystals formed where too small to be used on the single crystal X-ray diffractometer. Both the H-tube method, vapour diffusions and cooling were also attempted however no suitable crystals formed.

As the phenanthroline complex was proving difficult to get usable crystals, a database search was carried out to see if the triphenylphosphine complex had already been synthesised, to compare with the work undertaken by the former student. As found by that study only the triphenylphosphine oxide can be formed and that all of the halogen series had already been produced in this case[27][28][29] [18]. To still continue on from that work and another search on the Conquest database showed that there are many tetrahalotin complexes in which the tin coordinates to both a nitrogen and an oxygen atom such as picolinic acid and quinolone carboxylate[30][31]. So using the tetrabromotin similar complexes were made.

Reaction of tetrabromotin with 2,5-pyridinedicarboxylic acid and 4-(dimethylamino)pyridine

2,5-pyridinedicarboxylic acid (0.098g) and 4-(dimethylamino)pyridine (0.092g) were dissolved in a large sample vial in minimum methanol. SnBr4 (0.331g) was then added, along with the minimum methanol to ensure full dissolution. An attempt at recrystallization from methanol was made via slow evaporation over one week. This was unsuccessful as the fine white crystals formed where too small to be used on the single crystal X-ray diffractometer. Both the H-tube method, vapour diffusions and cooling were also attempted however no suitable crystals formed.

Reaction of tetrabromotin with Picolinic acid and 4-(dimethylamino)pyridine

Picolinic acid (0.150g) and 4-(dimethylamino)pyridine (0.096g) were dissolved in a large sample vial in minimum methanol. SnBr4 (0.333g) was then added, along with the minimum methanol to ensure full dissolution An attempt at recrystallization from methanol was made via slow evaporation over one week. This was unsuccessful as the fine white crystals formed where too small to be used on the single crystal X-ray diffractometer Both the H-tube method, vapour diffusions and cooling were also attempted however no suitable crystals formed.

Throughout this experiment a number of methods have been used in an attempt to grow suitable crystals, however in no case did the H-tube method work and evaporation yielded very fine powdery crystals for too small to be used with the single crystal X-ray diffractometer

Results

F

Cl

Br

I

N-Sn-N

76.5(3)

74.56(8)

-

72.91(9)

Sn-N

2.157(7)

2.234(2), 2.241(2)

-

2.292(3), 2.290(3)

Sn-X(trans to X)

1.887(5)

2.3779(8), 2.3970(8)

-

2.8195(3), 2.8213(3)

Sn-X(trans to N)

1.860(6)

2.3707(7), 2.3498(8)

-

2.7632(3). 2.7535(3)

Table 3. [SnX4(1,10-phenanthroline)]bond angles and lengths data. All lengths are shown in Angstrom (Ã…) and all angles in degrees.[18] [19] [25]

F

Cl

Br

I

N-Sn-N

75.1(1)

73.770(4)

74.085(1)

73.118(9)

Sn-N

2.181(3), 2.181(3)

2.224(6)

2.230(4)

2.277(7)

Sn-X(trans to X)

1.948(3), 1.940(3)

2.391(4)

2.560(9)

2.813(3)

Sn-X(trans to N)

1.924(3), 1.925(3)

2.371(5)

2.538(9)

2.787(1)

Table 4. [SnX4(2,2'-bipyridine)]bond angles and lengths data. All lengths are shown in Angstrom (Ã…) and all angles in degrees.[19][26][21]

Expected %

Actual %

Carbon

23.2

26.44

Hydrogen

1.3

1.53

Nitrogen

4.53

4.75

Bromine

51.68

-

Tin

19.2

-

R (Bromine and tin)

70.88

67.27

Table 5. CHN analysis results showing the composition of SnBr4(1,10-phenanthroline).

NMR analysis - See attached NMR of the sample obtained and of pure 1,10-phenanthroline for comparison.

Discussion

As can be seen from tables 3 and 4, the bond angles between the tin atom and the two nitrogen atoms gets smaller across the halogen series for both the 1,10-phenanthroline and the 2,2'-bipyridine complexes. This trend could be due to an increase in steric strain as going from fluorine through to iodine the size of the van der Waal’s radius of the halogen increases. The data also shows that for both series of complexes the tin-nitrogen bonds and the tin-halogen bonds lengthen as the halogen gets larger. This is due to the electronegativity of the halogen; the more electronegative the halogen, the shorter the bond due to the halogen pulling harder on the electron on the tin. This is consistent with the fluoride being the strongest Lewis acid of the four halides. In all of the chlorine, bromine and iodine complexes the Sn-X bonds are longer than the Sn-N bonds; however this is reversed in the case of fluorine. Again this could be due to the electronegativity of the respective atoms. Nitrogen has an electronegativity of 3.0 which is lower than that of fluorine so the fluorine will pull more on the tin shortening the bonds. This same reasoning can be applied to the bromine and iodine complexes as they have lower electronegativities, hence the shorter Sn-N bonds. Chlorine has the same electronegativity as nitrogen but has a Van der Waal’s radius of 20 pm larger than that of nitrogen so is more sterically hindered hence why although similar in electronegativity, the Sn-N bond is shorter.

The CHN analysis of SnBr4(1,10-phenanthroline) shown in table 4, shows good correlation with the expected results; any errors/deviations could be due to traces of solvent within the sample, or impurities from the synthesis such as contaminants on the glassware, unreacted starting materials, etc. The NMR is complex and has proved inconclusive, however a number of peaks can be seen that could possibly correspond to the hydrogen atoms on the phenanthroline. This could be due to a number of errors including a mixture of starting materials and product; incorrect preparation, instrument error, impurities in the sample, etc. There was insufficient time to carry out a repeat.

As crystals of SnBr4(1,10-phananthroline) were of insufficient size to use on the X-ray diffractometer, the series is incomplete. It is expected that as with the SnBr4(2,2’-bipyridine), that the complex should follow the trend and is an option for further study and development if a crystallisation method can be optimised and suitable crystals obtained. Although disappointing an inability to grow suitable crystals eaves option for further study and alternative methods and repeat attempts at growing such crystals could be made in order to complete the series such as using different solvents; using seed crystals and liquid-liquid diffusion.

The comparisons made here only apply to two tetrahalotin series so leaves option for further study to see if the same trends apply for a range of complexes. If this follows on, predictions could be made as to the properties of complexes in a series. Further study could also be carried out to look at the effects of electronegativity on the bond lengths, especially with a range of complexes with varying electronegativities of the coordinated atoms. Also the groups attached to the tetrahalotin molecules are fairly small in size so further study could look into larger coordinating groups to see if steric hindrance from them have an effect on the bond lengths or angles. If there is such an effect it would be predicted that this would influence the larger halogens more than the smaller halogens.

Taking all of the information obtained from both previous work and this project the paper for Acta Crystallographica has been put together and includes all of the refinement data, figures of the structures and comments on the structure.

Conclusions

This project has successfully found trends within two series of tetrahalotin complexes and links between the two, leaving options open for further study and investigation both within the phenanthroline series and alternative series. The trends show how simple fundamental properties, such as electronegativity and the size of atoms, has a measurable effect on the bonds within a larger complex. This information was used to put together a paper to be published in Acta Crystallographica by firstly refining the structure, analysing and comparing the data, producing suitable diagrams and writing up the journal article in the required style. The CHN analysis and the NMR (to a degree) are supporting evidence that the SnBr4(1,10-phenanthroline) synthesis was successful. This study also highlights the difficulties in growing crystals suitable for single crystal X-ray crystallography.