The Synthesis and Characterization of Ferrocene

A Modern Iterative Approach to a Classical Organometallic Laboratory Experiment

Pamela S. Tanner, Gennady Dantsin, Stephen M. Gross, Alistair J. Lees,
Clifford E. Myers, M. Stanley Whittingham and Wayne E. Jones, Jr. [1]

State University of New York at Binghamton, Binghamton, New York 13902

(Funded by the National Science Foundation)
(Submitted to J. Chemical Education)



Since ferrocene is credited with the rapid acceleration of modern organotransition metal chemistry (1,2) and the cyclopentadienyl group is extensively used as a stabilizing ligand, it is only fitting that the synthesis of ferrocene be incorporated into an advanced undergraduate inorganic laboratory. In our four credit course, the students work in pairs and have the opportunity to select six experiments from a total of nine. Three of these experiments must be selected from the area of materials chemistry and the topics include the synthesis of anhydrous CrCl3, a high temperature superconductor, the ZSM-5 zeolite and the lithium intercalation of WO3. Three wet experiments are also selected. These include the synthesis of W(CO)4, metal complexes of DMSO, a tris(bipyridyl)ruthenium complex, ferrocene, and the acetylation of ferrocene. If ferrocene is selected, it must be done in conjunction with the acetylation of ferrocene and these labs make up two of the three wet labs that are done by the student. Each lab incorporates an open ended question that the student may answer with the aid of library research or CAChe molecular modeling software with the Project Leader extension. This iterative approach builds confidence in the students ability to explore the unknown and reinforces the basic idea of the scientific method.

The ferrocene synthesis has been an extremely successful and popular selection. The students enjoy the diverse technical skills acquired during this experiment. These are techniques that a student may not be introduced to again as an undergraduate and include the use of air-less glassware while working on a vacuum line, cyclic voltammetry, bulk electrolysis, thin-layer and column chromatography. In addition, the compounds are characterized by standard methods such as melting point determination, IR and UV-Vis spectroscopies.


Experiments

Preparation of Ferrocene

Ferrocene is synthesized with a modification of the preparation reported by Jolly (3). The yield in the reported synthesis was 93% (3). Cyclopentadiene undergoes a 4+2 cycloaddition to form dicyclopentadiene. For this reason, cyclopentadiene is usually purified before use. Dicyclopentadiene boils at 170C and cyclopentadiene boils at 42.5 C. For efficiency, the dicyclopentadiene dimer is thermally cracked using a fractional distillation apparatus in advance by the teaching assistant. While this is usually done on the day of the experiment, we have found that cyclopentadiene may be stored without significant dimerization in a foil covered container in a freezer for several days. At the beginning of the lab period, the students grind KOH in a mortar and quickly transfer it to a tared vial. KOH is hygroscopic and should be ground in small portions (2 g). A nitrogen glove bag is a worthwhile investment for this step in the procedure. In addition to protecting the students from the corrosive KOH, it ensures that the KOH is dry. The FeCl2.4H20 will also go into solution more effectively if it is finely ground. It is then placed in a tared vial.
The pre-weighed KOH (15 g) is placed in a 100 mL (14/20) three-neck round bottom flask equipped with a magnetic stirring bar.

1,2-Dimethoxyethane (30 mL) is added with stirring to the KOH. One side of the neck is stoppered and the other is connected to a vacuum line through a gas adapter. While the mixture is slowly stirred and the flask is being purged with a stream of nitrogen, the cyclopentadiene (2.75 mL) is added. The resulting solution is rose colored. The main neck is then fitted with a pressure equalizing dropping funnel (25 mL) with its stopcock open. In a second one neck round bottom flask that is fitted with a septum, FeCl2.4H20 (3.25 g) and DMSO (12.5 mL) are stirred under a nitrogen atmosphere to dissolve the FeCl2.4H20.

After about five minutes, the stopcock is closed and the FeCl2 solution is added to the pressure equalizing dropping funnel. The reaction mixture in the three-neck flask is stirred vigorously and the purging with nitrogen is continued. After about ten minutes, the stopper is placed on the dropping funnel, the nitrogen flow is reduced and drop-by-drop addition of the FeCl2 solution is begun. The rate of addition is adjusted so that the entire solution is added in 30 minutes. Then the dropping funnel stopcock is closed and vigorous stirring of the dark green solution is continued for an additional 30 minutes. Finally, the nitrogen flow is stopped and the mixture is added to a mixture of 6M HCl (45 mL) and crushed ice (50 g). Some of the resulting slurry may be used to rinse the reaction flask to maximize the product yield. The slurry is stirred for about 15 minutes and the orange precipitate is collected on a Buchner or Hirsch funnel and washed with four 5-mL portions of water. The moist solid is spread out on a large watch glass and dried in the air. The compound is then purified through sublimation in a large glass petri dish that is placed on a warm hot plate (100 C). Care is used to avoid charring the ferrocene. The purified ferrocene is then characterized by melting point determination, UV-Vis and IR spectroscopies, and cyclic voltammetry. We are incorporating a bulk electrolysis to generate the ferrocenium cation.

Preparation of Acetylferrocene

Acetylferrocene is synthesized under mild conditions with a modification of the procedure reported by Bozak (4). The students are supplied with ferrocene during the second laboratory period so that the acetylation of ferrocene may take place concurrently with the purification of ferrocene. This encourages students to develop multi-tasking skills.

A mixture of ferrocene (1.5 g) and acetic anhydride (5 mL) is prepared in a small Erlenmeyer flask. To this mixture, 85% H3PO4 (1 mL) is added dropwise with constant stirring. This addition is exothermic and is accompanied by a change in color. Following the addition of the phosphoric acid, the Erlenmeyer flask is fitted with a CaCl2 drying tube. The dark green solution is then heated in a beaker of water on a hot plate for ten minutes (50 C). During this time, the solution becomes rose colored. The mixture is then poured over ice (20 g) into a large beaker that will accommodate the gas (CO2) formed during the NaHCO3 neutralization. Water is used to rinse the reaction flask and maximize the product yield. When the ice has melted, small quantities of sodium bicarbonate are added until gas evolution stops. The pH may be tested with pH paper to insure that neutrality is achieved. This is followed by cooling the resulting orange solution in an ice bath for 30 minutes during which time a brown precipitate forms. This precipitate is collected by suction filtration using a coarse fritted funnel. The dark brown solid is then washed with distilled water to remove impurities until it is pale orange in color. It is then dried in air for 15 minutes.

Thin layer chromatography is used to optimize the conditions for column chromatography of acetylferrocene. TLC plates (silica gel) are provided for student use. Alternatively, microscope slides may be used as TLC plates by applying a slurry that consists of silica gel (40 g) and chloroform (100 mL). A small amount of the crude acetylferrocene, which is a mono- and diacetylferrocene/ferrocene mixture, is dissolved in a vial in toluene (2-3 drops). A small amount of ferrocene is also dissolved in a separate vial in toluene. A line is penciled on each slide approximately 1 cm from the bottom of the TLC plate. The plates are spotted using a fine capillary applicator approximately on the pencil line. Each plate will contain two spots, one is ferrocene and one is crude acetylferrocene. The spots are allowed to air dry and then a second spot is applied at the same location to obtain a concentrated area of compound. The identity of the spot is indicated with a pencil mark. The plates are individually placed with the spotted end in the solvent in five developing chambers. The chambers contain the following solutions: petroleum ether, toluene, ethyl ether, ethyl acetate and a mixture of 10% ethyl acetate and 90% petroleum ether. The pencil mark should be above the solvent level. The solvent containers are covered while the plates are developing. The plates are removed when the solvent front has traveled approximately 3/4 of the distance of the plate. The plates are air dried. The TLC plates may be developed in an iodine chamber. This will result in brown spots that can be marked and identified so that the plates may be included in a laboratory report. The solutions that provide maximum separation of the two components are chosen as column chromatography solutions. For instance, ferrocene may elute with toluene while the acetylferrocene remains on the column and is then eluted with a toluene/ethyl acetate mixture. The color of the spots is helpful to discern the individual bands that elute from the column. The crude acetylferrocene is dissolved in the solution that is selected to elute the first component.

The column is assembled by placing a small piece of glass wool into the bottom of the column (50 mL buret). The glass wool is then covered with a small amount of sand and the buret is filled with the solvent that was chosen to dissolve the crude mixture. A powder funnel is used to slowly fill the column with dry silica gel to a height of approximately 30 cm. The column is never allowed to dry. Alternately, the column may be prepared by the traditional slurry method. A small amount of silica gel may be added to the crude acetylferrocene solution to make a slurry that is then added to the top of the column and covered with a small amount of sand. The two solutions (or mixtures) are then used to purify the crude acetylferrocene. The ferrocene band is discarded and the solvent is removed from the acetylferrocene band by rotary evaporation. It may then be recrystallized from chloroform. The acetylferrocene is characterized by melting point determination, IR and UV-Vis spectroscopies, and cyclic voltammetry.


Discussion

The experimental procedure for the synthesis of ferrocene provided above was adopted after several failed attempts to incorporate newer microscale techniques that utilize ethylene glycol (5) as the solvent rather than 1,2-dimethoxyethane. When ethylene glycol was used, an extremely viscous reaction mixture resulted that was incapable of being stirred effectively in the micro-glassware. Our success rate with the revised preparation is 100%.

Our advanced undergraduate inorganic lab is taught in the semester format with two three-hour weekly classes. The students learn to multi-task to accomplish their lab responsibilities efficiently. We have provided the following suggested format (Table 1) to accomplish the synthesis and characterization of ferrocene and acetylferrocene in two and a half weeks. This format is not provided to the students. They are innovative and are required to submit their own schedules before beginning work. The format allows instructors and teaching assistants to flexibility in the method of ensuring that the students use their time efficiently.


Table 1. Suggested Time Management Schedule
DayProgram
1Synthesis of Cp2Fe; teaching assistant to provide cyclopentadiene
2Sublimation of Cp2Fe; students are given Cp2Fe to perform the acetylation
3Thin layer and column chromatography of acetylferrocene followed by rotary evaporation; begin characterization of Cp2Fe (melting point, UV-Vis, IR)
4Characterization of acetylferrocene (melting point, UV-Vis, IR); CAChe modeling
5Finish characterization including cyclic voltammetry and bulk electrolysis


Crude ferrocene and acetylferrocene were synthesized in 51-79% and 27-58% yield respectively. An experimental melting point range of 169-171 C was obtained for ferrocene. The reported melting point range is 173-174 C (3). For acetylferrocene, the experimental melting point range was 80-83 C as compared with the reported range of 81-83 C (7). Infrared spectroscopy was performed by the students on ferrocene and acetylferrocene both as a KBr pellet and as a Nujol mull on NaCl plates. The infrared spectra were comparable to those reported for ferrocene (3) and acetylferrocene (8). The main difference between the spectra of ferrocene and acetylferrocene is of course the appearance of a carbonyl stretch at 1736 cm-1 that is present in the acetylferrocene and absent in the ferrocene. Some students also observed a peak at 893 cm-1 that is attributed to the monoacetylferrocene ring. They did not observe peaks that could be attributed to the 1,2-diacetylferrocene complex at 917 cm-1 or a doublet due to the 1,3-diacetylferrocene complex at 922 and 905 cm-1 (8). The experimental UV-Vis spectra of ferrocene and acetylferrocene were obtained in acetonitrile and Beers law was used to calculate the molar absorptivity. The UV spectrum for ferrocene shows maxima at 330 nm (2 = 52) and 440 nm (2 = 90), and a rising short-wavelength absorption at 225 nm (2 = 5051). This is comparable with the reported spectrum in ethanol (3). The UV spectrum for acetylferrocene shows maxima at 219 nm (2 = 2.2 x 104), 266 nm (2 = 5268) and 320 nm (2 = 1124). Except for the calculated molar absorptivity of the peak at 219 nm, this is comparable with the reported spectrum in 95% ethanol (8). The students also observed peaks assigned to ferrocene in their acetylferrocene samples.

The electrochemistry component of this laboratory was the first time that most students were exposed to cyclic voltammetry and the bulk electrolysis technique. An Amel System 5000 Potentiostat was used for all measurements. For cyclic voltammetry, the electrochemical cell was a 100 mL beaker equipped with a Ag/AgCl reference electrode (student prepared), a BAS (West Lafayette, IN) platinum-disk working electrode (2 mm diameter) and a large (1 cm2) platinum flag counter electrode. After having verified a flat background of tetrabutylammonium hexafluorophosphate (0.01 M) supporting electrolyte in acetonitrile in the range 0.0 to 1.0 V vs. Ag/AgCl, cyclic voltammograms of ferrocene and acetylferrocene (approximately 3.2 x 10-3 M) were obtained at scan rates of 100 500 mV/sec. A typical cyclic voltammogram of ferrocene showed a reversible oxidation at E1/2 = +0.35 V vs. Ag/AgCl with Ep/2 = 0.057V. A typical cyclic voltammogram of acetylferrocene also showed a reversible oxidation at E1/2 = +0.58 V vs. Ag/AgCl with Ep/2 = 0.044V. Small peaks for ferrocene were also visible in the acetylferrocene cyclic voltammogram. These results are comparable to the reported E of acetylferrocene at +0.27 V vs. the ferrocene/ferrocenium couple (6).

A second new electrochemical component that was recently introduced into this laboratory is the bulk electrolysis of ferrocene to ferrocenium. The electrochemical cell was a 100 mL beaker equipped with an Ag/AgCl reference electrode (student prepared), a BAS (West Lafayette, IN) reticulated vitreous carbon (RVC) working electrode and an extremely large platinum flag counter electrode. After having verified a flat background of tetrabutylammonium hexafluorophosphate (0.01 M) supporting electrolyte in acetonitrile in the range 0.0 to 1.0 V vs. Ag/AgCl, the bulk electrolysis of ferrocene (approximately 7.5 x 10-4 M) was achieved on several occasions. As expected, a new peak in the UV-Vis was observed at 620 nm and the solution changed color from orange to blue. Unfortunately to date, these experimental conditions are not reproducible.

As a supplement to their standard chemical characterization, students used the CAChe molecular modeling program to build a ferrocene molecule in both the eclipsed and staggered conformations and to remove an electron to obtain information about the ferrocenium cation. The results of this modeling were then discussed in relation to their experimental observations.

When the students have synthesized and derivatized ferrocene, they have an experimental background for comparison of the unsubstituted ferrocene versus the acetylated ferrocene. They also have a clear understanding of the potential R groups that are chemically practical. This is especially meaningful if the student has completed organic chemistry and is able to relate the familiar benzene substituents with the ferrocene molecule. We have found that if a student proceeds through the iterative question before understanding the acetylation experiment, they design strange, wondrous and impractical molecules with the aid of the CAChe system. It must be stressed that molecular modeling is only a tool. The input is influenced to a large degree by the understanding of the operator which may be enhanced with guidance from the instructor.

A natural progression at the completion of the two syntheses is the introduction of the iterative question. Students are asked to design a ferrocene with specific properties such as a different colored ferrocene. This question is answered with the aid of CAChe modeling where electronic spectra of the gas phase ferrocene and the substituted ferrocene may be generated by ZINDO (Zerners Intermediate Neglect of Differential Overlap). A more comprehensive iterative project involves both library work and molecular modeling. The students are asked to find the preparation of a substituted ferrocene in the library. They may also design a synthesis and confirm the synthesis with the aid of library references. They then model the complex and predict its spectroscopic characteristics based upon what they are able to calculate from the molecular model and their knowledge of general chemical trends.

Since the students became familiar with cyclic voltammetry, one trend of interest involves the ionization potential of the substituted ferrocenes. One student project involved a comparison of several known substituted ferrocenes (6) and their gas phase models (Figure 1 and Table 2). The gas phase models were used since the expected solvent dependence has not been observed using the CAChe system due to initial limitations with project leader. The initial calculated ionization potentials were adjusted by subtracting 7.647 eV. This sets the ferrocene/ferrocenium couple at zero as is customary (6). These values and a least squares regression plot were then plotted. In general, a downward trend in the least squares regression is observed with the more easily reduced ferrocenes containing electron withdrawing substituents having positive ionization potentials. Conversely, the more easily oxidized ferrocenes with electron donating substituents are calculated with negative ionization potentials. Deviations from experimental data may be accounted for since the student was comparing gas phase ferrocene models and acetonitrile ferrocene electrochemistry (6).


Fig. 1. Student CAChe Project


Table 2 Student CAChe Project


Conclusion

The incorporation of an iterative question into each of our advanced inorganic undergraduate laboratories has allowed students to plumb the depths of their chemical knowledge and to acquire new tools that improve their use of the scientific method. The students enjoy the high success rate of the ferrocene/acetylferrocene lab. They also appreciate the chance to acquire new synthetic techniques such as the use of Schlenk techniques. In addition, the use of novel instrumental analysis such as electrochemistry is beneficial to their overall undergraduate education. They seem to thrive on the diverse exposure and the opportunity to stretch themselves. This allows them to become excited about chemistry and like the experiment that they are conducting, they come full circle and view chemistry in a new light as a useful, valuable tool. The addition of the iterative question to a classical laboratory can therefore provide an additional richness to the traditional wet chemistry.


Acknowledgments

Research supported by NSF under Grants # DUE-9452023 and DUE-9452131.


Literature Cited

1. Kauffman, George B. J. Chem. Educ. 1983, 60, 185.
2. Kealy, T. J.; Pauson, P. L. Nature 1951, 168, 1039.
3. Jolly, W. L., The Synthesis and Characterization of Inorganic Compounds, Prentice-Hall: New Jersey, 1970.
4. Bozak, R. E. J. Chem. Educ. 1966, 43, 73.
5. Szafran, Z.; Pike, R. M.; Singh, M. M., Microscale Inorganic Chemistry, Wiley: New York, 1991.
6. Geiger, William E. J. Organomet. Chem. 1990, 22, 142.
7. Wade, Leroy G. J. Chem. Educ. 1978, 55, 208.
8. Rosenblum, Myron, Chemistry of the Iron Group Metallocenes: Ferrocene, Ruthenocene, Osmocene Part One, Interscience Publishers: New York, 1965.


Footnote

[1] Pamela S. Tanner, Wayne E. Jones, Jr., Clifford E. Myers and M. Stanley Whittingham. Presented at the 210th National Meeting of the American Chemical Society, Chicago, IL, August 1995; paper CHED 51.


Copyright © 1989-1995 The Research Foundation of SUNY, et al