Urea substituted phosporamidite ligand complex as a
catalyst in allylic substitution reactions
4 april 2013 – 26 july 2013
HomKat – Homogenous and Supramolecular Catalysis
Rosa Kromhout
10003493
Supervisors:
Yasemin Gümrükçü
Date:
Prof. dr. Joost H. Reek
31 Januari 2014
Second reviewer:
Dr. N.R. Shiju
CONTENTS CHAPTER PAGE 1 Abstract 3 2 Populaire samenvatting 4 3 Introduction 5 4 Goal of research 8
5 Results and discussion 9
6 Conclusion 17
7 Prospects / Further investigation 17
8 Experimental data 18
1 ABSTRACT
Transition metal-catalyzed nucleophilic allylic substitutions reactions are useful methods for carbon-carbon and carbon-carbon-heteroatom bond formation. The direct substitution of allylic alcohols results in only water as a byproduct and is a progress for the more atom economical and environmentally friendly transformations. Recently, a phosphoramidite palladium complex is developed as a catalyst for this type of reactions. It has been found that an addition of catalytic amount of urea during the catalysis improves the activity. A possible explanation for this improvement was that the hydrogen bond interaction between allylic alcohol and the catalytic amount of urea additive assists in the activation of the alcohol. The leaving ability of the hydroxyl group is promoted, resulting in better conversions when compared to catalysis without additional urea. This interesting discovery raises the question what role urea plays when it is substituted in a ligand. For this research the goal is to determine the effect is of urea substituted phosphoramidite ligand in catalysis of direct activation of allylic alcohols.
The new phosphoramidite ligand with an urea substituent is successfully synthesized and analyzed with spectroscopic techniques. A palladium catalyst that is formed with this ligand is tested for amination reaction of cinnamyl alcohol with N-methylaniline. The same reaction was pursued for kinetic studies. The results of the kinectic studies were compared with previous catalyst that was using urea as an additive. Also commercially available monophos, that has no possibility to form any type of hydrogen bond with the substrate, was tested for this reaction. Based on those results we conclude that the new complex is only active at 80⁰C where others are at room temperature and a catalytic additional amount of free urea still assists in the catalysis. Furthermore is discovered that the chirality of ligands also plays a role in the activity, chiral ligands carry out the reaction with higher conversions. The lack of a chiral center in the new ligand could be one of the reasons of low activity.
2 POPULAIRE SAMENVATTING
De overgangsmetaal gekatalyseerde nucleofiele allylische substitutie reactie is een handige en effectieve manier om koolstof-koolstof of koolstof-hetero atoom verbindingen te maken. Allylische alcoholen kunnen zonder voorbewerking van de hydroxygroep gekatalyseerd worden. De grote voordelen hiervan zijn is dat de stoichiometrische hoeveelheden afval van de voorbewerking wegvalt en alleen licht milieu belastende water wordt geproduceerd als bijproduct. Onlangs is er een fosforamidiet-paladiumcomplex gesynthetiseerd dat deze reactie kan katalyseren. Ontdekt is dat de waterstofbruginteractie tussen de alcohol en additionele urea voor een betere conversie tijdens de katalyse zorgt in vergelijking met katalyses zonder additionele urea. Een mogelijke verklaring is dat deze waterstofbrug de vertrekmogelijkheid van de hydroxygroep bevorderd. Deze waarneming leidt tot de vraag wat voor effect een urea heeft wanneer deze is ingevoegd in het ligand. Het doel van dit onderzoek is onderzoeken wat de effecten zijn van een urea gesubstituteerd ligand in de directe katalysse.
Het nieuwe ligand is succesvol gesynthetiseerd en geanalyseerd met spectroscopische technieken. Met het palldiumcomplex zijn katalytische reacties tussen cinnamyl alcohol en N-methylaniline uitgevoerd en de kinetiek van de reacties is bestudeerd. De verkregen resultaten zijn vergeleken met de resultaten van afgelopen onderzoeken waarin urea is berbuikt als additief. Een monophos ligand, dat geen waterstofbrug kan vormen met het substraat, is ook onderzocht. Uit de resultaten bleek dat het nieuwe palladiumcomplax actiever is op 80⁰C, waar andere katalysotoren actief zijn op kamer temperatuur en dat een additieve hoeveelheid urea nog steeds assisteert tijdens de katalytische reactie. Daarnaast bleek chiraliteit van het ligand ook invloed te hebben op de activiteit van een katalysator. Een katalysator dat chiraliteit bevat voert dezelfde reactie met hogere conversies uit. Het urea gesubstitueerde ligand mist deze chiraliteit, wat een van de oorzaken kan zijn waardoor de conversie lager is dan de eerder onderzochte liganden.
3 INTRODUCTION
In the last century the modern pharmaceutical industries have made remarkable achievements in the organic chemistry. However, most of the unveiled syntheses were developed without taking the sustainability and waste ratio in consideration. The discovery of the ability of transition metal complexes to lower the energy barrier of reactions has played a major part in the improvements. In the last few decades many other improvements were made on the field of sustainability and atom economy. Reactions which at first were unable to be performed under extreme reaction conditions, can now be carried out at mild reaction conditions. Many reactions were reviewed again and with the developments new catalytic pathways were discovered.
In this research we focused on the nucleophilic substitution reactions, that are used in formation of carbon-carbon and carbon-heteroatom bonds.1 Using allylic alcohols for this reaction is interesting
for the cleaner process. However, the hydroxyl group has poor leaving abilities and it needs to be activated in order to pursue the catalysis. One of the solutions is the substitution of allylic alcohols with good leaving groups, such as halogens, esters or phosphates.2 The stoichiometric amounts of
waste was formed end of the reaction. The pre-activation step and the formation of stoichiometric amounts of waste were the major disadvantages of this type of nucleophilic reaction. (Figure 1)
Figure 1: Two reaction paths for the nucleophilic allylic substitution reaction. The upper reaction pathway requires the pre-activation step of the alcohol, afterwards the nucleophilic attack takes place. The lower reaction pathway describes
the direct activation of allylic alcohol, were water is the only side product.
However, the nucleophilic substitution reaction can be a very promising reaction when the allylic alcohol is directly used in catalysis without pre-activation.3 This will eliminate the stoichiometric
amounts of waste and water is the only side product. Water has a lower hazard quotient (Q) than halides and phosphates, which makes it a more environment friendly waste product.4 The hazard
quotient indicates quantitatively what is the impact of the type of compound on the environment. Absolute Q-values are difficult to assign, because the effects of waste differ according to location and type. Also, with the direct catalysis the E-factor, kg of waste over kg of produced product, is lower than the reactions with the pre-activation step. By multiplying Q and the E-factor a better measure is given of the impact on the environment than these factors alone. The EQ of water as a sole waste product is significantly lower compared to the EQ-value of the stoichiometric amounts of waste. Researchers have performed the direct amination of allylic alcohols with various transition metal catalysts. Ruthenium, Iridium and platinum catalysts were developed by various research groups.5, 7
Though, palladium is the most used transition metal used in allylic amination substitution reactions. Since 1964 various palladium complexes were reported, these complexes often contained small
ligands which had no particular participation in the selectivity of the substitution.6 To increase the
selectivity, and also the activity, additional additives were added. Researchers have reported organic additives, such as CO2 gasses under high pressure, and inorganic additives, such as transition metal
oxides.7 These additives were added in catalytic or stoichiometric amounts to activate the C-O bond
and promote the leaving ability of the hydroxyl group.
Unfortunately these additives make the entire substitution reaction less atom economical. However, in the last decade, several palladium catalyst were reported which perform the allylic substitution reaction without any additives. Mono- or bidentate phosphine ligands were used in the catalysis of amination of allylic alcohols. The research groups of Le Floch8 and Breit9 have studied a range of
palladium complexes for the amination of different types of allylic alcohols. These catalysts can activate the allylic alcohol directly and that increases the atom economy. Therefore, palladium based catalyst are of interest in this study.
The catalysts studied by Le Floch and Breit often involved, next to the ligands, a η3-π-allyl compound,
which structure is very similar to allylic alcohols. During the catalysis the η3-π-allyl compound gets
replaced by the allylic alcohol. The catalyst promotes the hydroxyl group leaving ability of the substrate and a nucleophile will be able to attack. An attack of the nucleophile can result in two isomeric products, one linear product and one branched product, and the sole byproduct, water. (Figure 2)
Figure 2: a. Palladium metal complex b. Reaction scheme of nucleophilic allylic substitution reaction with intermediate
Remarkable, ligands that contain phosphor, sulfur or nitrogen have the most influence on the conversion in the substitution reactions. Most likely this is caused by their property to donate electrons into the orbitals of the transition metals. The high electron density on the metal is probably needed for the stabilization of the substrate-catalyst intermediate. When a free allyl coordinates with the catalyst there forms a sigma donation from the π system at the allyl to the pz orbital metal
or other suitable orbitals.10 (Figure 3) This is the lowest π orbital energy level, the next orbital is a
non-bonding orbital. Depending to the electron distribution the allyl will act as a donor or acceptor, mostly this is a filled donor orbital. The suitable orbitals of the metal are py and dyz orbitals. The last
and highest orbital acts as a acceptor for the π-back donation from the dxz op px orbital of the metal.
If a substrate coordinates the electron distribution will likely be different such that the non-bonding is not a donor but an acceptor. The hydroxyl group on the substrate is an electron withdrawing group, causing an empty non-bonding orbital. The electron density on the metal is increased by the electron donating ligands and probably donates electrons into the non-bonding orbital of the allyl. These electron exchanges cause the coordination and stabilization of the catalyst-substrate complexes. With various ligands various electronic properties can be obtained.
Figure 3: Coordination of η3-allyl compound and a metal by sigma donation and π-back donation.10
In previous studies various ligands were screened for their selectivity in amination reactions as shown in figure 2b. The used reaction during the screening is the allylic amination reaction of cinnamyl alcohol and N-methylaniline. (Figure 4) Discovered was that phosophoramidite ligands have great selectivity towards the linear product. The best performing phosphoramidite (1) contained an amino acid attached to the backbone. (Figure 5)
Figure 4: Amination reaction with cinnamyl alcohol and N-methylaniline
Figure 5: a. Ligand 1 with descriptions of the backbone and amino acid section b. Complex structure of a 2:1 ligand: palladium ratio.
During the same screening process also is discovered that an addition of a catalytic amount of urea improves the activity of the catalyst in amination of allylic alcohols. Most likely it is the effect of hydrogen bond formation between the hydroxyl group and the urea. This bond formation is helping to activate the C-O bond and promoting the leaving ability. (Figure 6) However, the addition of urea makes the overall substitution reaction less atom economical. Therefore, it is interesting to investigate whether the urea substituted ligand has any effect on the selectivity and activity of the catalyst which is applied for the substitution reaction. The main idea was that the inserted urea to the phosphoramidite ligand will increase the selectivity compared to ligands without urea inserted. In this case the urea substituent might be closer to the hydroxyl group and therefore this ligand might help to increase the activity of the catalyst compared to non-substituted phosphotamidite ligands.
First, we prepared a new urea substituted phosphoramitide ligand and analyzed its structure with analytical spectroscopic measurements. Then the palladium complex of this ligand is prepared with (Pd-allyl(COD))BF4 precursor. Kinetic studies were performed in presence and absence of additional
free urea to determine whether the free urea effects the results of the catalysis.
4 GOAL OF RESEARCH
The goal is to determine the influence of urea substituted phosphoramidite ligand complexes in the allylic nucleophilic substitution reaction of cinnamyl alcohol and N-methylaniline and compare the results to earlier studied ligands.
5 RESULTS AND DISCUSSION
Ligand synthesis
The synthesis of urea substituted ligand, 2, was completed in three steps.
Figure 7: Ligand 2
First the backbone of the ligand is synthesized. This diol is synthesized with a commercial available alcohol, 3-tert-butyl-4-hydroxyanisole, potassium ferricyanide (K3Fe(CN)6) and KOH. (Figure 8)
Figure 8: Reaction scheme of the synthesis of the diol backbone
The urea part of the ligand was synthesized separately with the straight formed reaction of commercial available ethylenediamine and phenyl isocyanate. (Figure 9)
Figure 9: Reaction scheme of the synthesis of the urea amine
The P-Cl backbone of 2 was synthesized with the diol and PCl3. The addition of urea amine resulted in
the phosphoramidite ligand (2) successfully.
Figure 10: Reaction scheme of the synthesis of ligand 2
The characterization of ligand 2 was done by 1H-NMR is reported here for the first time. The peaks
Figure 11: Ligand 2
Figure 12: 1H-NMR spectrum of the ligand 2
Table 1: Data set of 1H NMR (300 MHz, CD
2Cl2) from ligand 2
Signal (ppm) Data Signal (ppm) Data
7.27 (m, 5H) 3.82 (s, 6H)
6.99 (d, J = 3.1 Hz, 2H) 3.44 (dt, J = 29.4, 6.6 Hz, 1H)
6.71 (d, J = 3.0 Hz, 2H) 3.19 (q, J = 5.9 Hz, 2H)
6.24 (s, 1H) 2.89 (m, 2H)
4.83 (t, 1H) 1.49 (s, 18H)
The peaks at 7,27 ppm have an integral of 5H and are in the typical range for phenyl protons. These peaks correspond with the protons of number 10. Signals 6,99, 6,71, 3,82 and 1,49 are identical to the signals from the 1H NMR of the backbone with the correct integrals. (Exprimental data, figure 18)
Figure 13: 2D cosy spectrum of ligand 2
The correlation of peaks at 6,99 and 6,71 ppm confirms that these peaks are corresponding with protons of number 3 and 1 on the backbone. The proton of the peak at 6,24 ppm gives a broad N-H signal and has no correlations, confirming the assignment to proton 9. The peak at 3,44 ppm correlates with the peak at 2,89 ppm and is a doublet of triplets and has an integral of 1H. A doublet of triplets means that a proton has two neighboring proton groups, one gives a doublet and the other a triplet. Only one of the remaining protons can have such a multiplet and that is proton number 8, the protons at 7 gives a triplet and the long distance coupling with 9 gives the doublet. In a 2D cosy measurement long distance couplings are not visible. Proton 7 (2,89 ppm) also correlates with the peak at 3,19 ppm, which is proton number 6. The last correlation is the 6, 5 correlation, confirming the peak at 4,83 is proton number 5.
The 31P-NMR spectrum shows a peak at 145 ppm, corresponding with the phosphorus atom of the
Figure 14: 31P NMR spectrum of ligand 2
Complex formation
After the synthesis and purification of the ligand we formed the Palladium complexes with [(η3
-allyl)Pd(cod)]BF4 precursor. Two equivalents of ligand 2 were used and a structure prediction was
made as shown in figure 13.
The 31P-NMR measurements showed a shift in phosphorous peaks for 2. The peak at 145
disappeared, concluding all free ligand coordinated to the palladium precursor. At 132 ppm a broad peak appeared and variable-temperature 31P-NMR measurements were made for further
investigation for this peak. (Figure 16) The VT-diagram showed splitting and sharpening of the peaks at 132 ppm when the temperature was lowered. Indicating various conformations of the complex is present. The broad peak is that is observed at room temperature should be the complexes that are in exchange. Cooling down slows down the exchange and helps to identify these complexes. All the coupling constants are the same and four sets of doublets were visible in the VT-diagram. The big peak at 132 remains visible, which might represent a different configuration of the complex, possibly the coordination of de ligands is different than the structure corresponding with the other isomers. The structures of the complexes were not further investigated, due to lack of time and limited available amounts of ligand 2.
Figure 16: VT-diagram of the P-NMRs of the complex. Lowest spectrum is measured at 23⁰C and the highest at -70⁰C.
Catalytic reaction and kinetics
In order to investigate the effect of this new ligand 2 in an allylic amination reaction, we followed the reaction in time. Two parallel experiments were done in the presence and absence of urea for the kinetics analysis. (Reaction as shown in figure 4) The conversions were calculated according the internal standard, 1,3,5-trimethoxybenzene in the GC. Dimethylfumarate was added to the samples are taken from the reaction mixture to prevent any further conversion when injected into a GC. Here we analyzed the GC data of catalyst with this new ligand 2 (cat 2). Later on cat 2 is compared with the previous studied catalyst with ligand 1 (cat 1). The data points are shown in graph 1.
Graph 1: The conversion vs time graph of catalyst 2 in presence and absence of urea at 80⁰C.
The reaction profile after heating to 80⁰C is demonstrated in graph 1. The pre-stirring at room temperature showed only traces of activity, thus the reactions mixtures were heated up to 80oC. The
pre-stirring at r.t. might have caused some deactivation of the catalyst. However, the fast response of the catalyst to the heating proofs that the catalyst was still active. The yield was 69% in presence of urea additive, whereas it was 28% in absence of it. A 1H-NMR study of a catalytic reaction which
directly started at 80oC and in absence of urea confirmed that the catalyst was indeed significantly
more active compared to a pre-stirred catalysis. According to that 1H-NMR study a yield of about 70%
was obtained at 80oC in 11 hours in absence of free additional urea. See the experimental data
section for the 1H-NMR spectrum. Unfortunately, not enough amounts of ligand 2 was left over to
perform another catalytic experiment. Nevertheless, we compared the data of graph 1 with earlier obtained data of the catalyst with ligand 1 (cat 1) and a commercial available ligand (cat 3).
The data of catalyst 2 were compared with the data of catalyst 1 (Graph 2) and with the data of chiral R-monophos complexes (cat 3). (Graph 3)
Graph 2: The comparison of the graphs of catalyst 1 and catalyst 2 in presence and absence of urea
The data points of catalyst 1 were reported earlier and used here for comparison. The kinetic studies of catalyst 1 were done at r.t. However, the reactions were done at 80oC for catalyst 2 due to its poor
activity. Changing the reaction temperature for catalyst 1 could be an option for a fair comparison, but catalyst 1 was fairly fast at 80oC, that makes the kinetic study inconvenient. Catalyst 1 resulted in
yields of 92% and 73% in presence and absence of urea, respectively, after 4 hours. Whereas ligand 2
0 0,05 0,1 0,15 0 5 10 15 20 C o n ce n tra ti o n (M ) Time (h)
With additional urea Without additional urea
0 0,05 0,1 0,15 0,2 0,25 0 5 10 15 20 C o n ce n tra ti o n (M ) Time (h) 1 1 + urea 2 2 + urea
Graph 3: The comparison of graphs of catalyst 3 in presence and absence Figure 17: R-monophos of urea and catalyst 1 and 2.
Graph 3 shows the profile of catalyst 3, with chiral R-monophos as ligand (3, figure 17), in presence and absence of additional urea at room temperature. Chiral R-monophos is commercially available. Graph 3 also shows the profiles of catalyst 1 and 2. The kinetic study with catalyst 3 in presence of additional urea doesn’t show any difference in the catalysis in absence of it. Concluding that the catalysis is not affected by the hydrogen bond formation between the hydroxyl group of the allylic alcohol and urea. Catalyst 3 probably doesn’t allow to form the hydrogen bond between the allylic alcohol and the free urea. This is not further investigated. Catalyst 3 performs the catalysis in a yield of 44% after four hours. Which is better than catalyst 2 after an incubation time of 17 hours at r.t. and 4 hours of reaction time at 80oC. Though, catalyst 1 has a better conversion than catalyst 3,
meaning it has a better activity in this type of amination reactions.
Graph 4: The conversion vs time graph of catalyst 3 and catalyst 4 in presence and absence of urea
Ligands 1 and 2 differ also in chirality. In order to determine whether chirality of the ligand indeed has effect on the activity, a kinetic study was conducted with a catalyst with a racemic mixture of ligands R and S monophos and the same palladium precursor (cat 4). In comparison with catalyst 3 shows that the conversion of catalyst 4 is lower. The free urea does have an effect on the activity and therefore the conversion of catalyst 4. The yields were 28% and 37% after 5 hours of reaction time, which is significantly lower than catalyst 3 after 5 hours. Concluded is that chirality of ligands has influence on this catalytic amination reaction.
0 0,05 0,1 0,15 0,2 0,25 0 5 10 15 20 C o n ce n tra ti o n (M ) Time (h)
1 1 + urea 2 2 + urea 3 3 + urea
0 0,05 0,1 0,15 0,2 0 2 4 6 8 10 C o n ce n tra ti o n (M ) Time (h) 3 3 + urea 4 4 + urea
Chirality of the catalyst
With the kinetic studies is discovered that chirality of the ligand has a role in the catalysis. Chiral molecules tend to circularly polarize absorbed linearly polarized light. A circular diochroism (CD) measurement shows the sum of the detected left and right circularly polarized light. The spectropolarimeter (CD instrument) detects the two types of circularly polarized light separately and displays at a given wavelength its diochroism.11 The difference between the amounts of each
circularly polarized light causes a peak in the diochroism, this is also called the cotton effect.
Both the R and S configurations of the monophos were investigated. Since these compounds are chiral opposites of each other, expected is that the diochroism will be each other’s opposites as well.
Graph 5: CD spectra of chiral R- and S-monophos
As seen in graph 5 chiral monophos ligands R and S polarize incoming light contrary as expected. In the range between 230 nm and 340 nm either the left or the right polarized light is in excesses compared to the opposite type of polarized light. From 340 nm silence is observed, this is caused by the absorbance region of monophos, which is between 280-340 nm. Linear polarized light outside this region will not be circularly polarized, since it will never be absorbed by monophos.
Graph 6: CD spectra of ligand 1 and the amino acid phenylanaline
The spectra of ligand 1 and the free amino acid used in the synthesis of ligand 1 were measured.
-600 -400 -200 0 200 400 600 180 230 280 330 380 430 Wavelength (nm) S-Monophos R-Monophos -150 -100 -50 0 50 100 190 240 290 340 390 Wavelength (nm) Ligand 1 Phenylanaline
Graph 7: CD spectra of ligand 2 compared with the achiral methanol
Ligand 2 has no chiral center, thus the diochroism is expected to show silence. The spectrum was compared to another non-chiral compound, methanol. (Graph 7) The diochroism of ligand 2 was almost identically confirming the expectations. The peaks in the low wavelength regions are noise signals.
6 CONCLUSIONS
In conclusion, the urea substituted phosphoramidite ligand was successfully synthesized and with spectroscopic measurements the structure is confirmed. The catalyst with this ligand and Pd-precursors show activity in the direct activation of allylic alcohols in amination reactions at 80⁰C. The selectivity of this ligand is similar compared to ligand 1, only the linear product is formed. However, less activity was observed compared to ligand 1, both in presence and absence of additional urea. A chiral substituent to the backbone makes the backbone also chiral. This increase of chirality could be one of the reasons why ligand 1 performed better in this catalysis. We did not investigate the exact structure of the new catalyst and reason why its activity is lower.
7 PROSPECTS / FURTHER INVERSTIGATION
Additional spectroscopic measurements are needed for the characterization of the new complex, such as mass spectroscopy and x-ray diffraction.
Additional kinetic studies should be performed with the new complex. Due to complications in our first try, the new kinetics should be started at 80oC for fair comparison.
New complexes can be formed with ligand 2, with varying ligand palladium ratios or with different palladium precursors.
The electronic properties of urea can be an influence on the hydrogen bonds formation. Various urea substituted ligands could be synthesized and analyzed. Since the chirality plays a role in the catalysis, I would suggest synthesizing ligands with chiral urea substituents.
-60 -40 -20 0 20 40 180 280 380 480 Wavelength (nm) MeOH Ligand 2
8 EXPERIMENTAL DATA
All reactions and experiments were carried out and under inert atmospheres in dry solvents with syringe and standard schlenk techniques. All glassware was oven dried or flame dried. All reagents were obtained from commercial suppliers and used without further purification, unless stated otherwise. 1H and 31P NMR spectra were measured with Bruker spectrometer (1H: 300 MHz and 400
MHz; 31P: 121 MHz). GC analysis of the kinetic studies was performed with a Interscience Focus Chiral
GC with varian capillary column (CP-Chirasil-Dex CB 25 m column, 0,32 mm diameter, 0,25 μm film thickness).
Synthesis of 5,5’-dimethoxy-3,3’-di tert-butylbiphenyl-2,2’-diol12
A solution of 3-ter-butyl-4-hydroxyanisole (4,00 g, 22,2 mmol) in methanol (120 mL) was prepared, and a solution of KOH (4,43 g, 79,2 mmol) and K3Fe(CN)6 (7,32 g, 22,2
mmol) in water (120 mL) was added dropwise over 1 h at room temperature. The mixture was stirred for 2 h before the addition of 80 mL of water. The suspension was extracted with 200 mL of ethyl acetate twice. The aqueous solution was extracted with 60 mL of ether, and the organic phases were combined and washed with 80 mL of saturated brine. The organic phase was dried over Mg2SO4. Removal of the solvents resulted in a
brown solid. Washing with n-hexane resulted in light brown powder; yields of several batches: 3,61 g to 3,94 g (90% to 98%)1H NMR (400 MHz, CDCl
3) δ 7.28 (s, 4H), 6.99 (d, J = 3.0 Hz, 2H), 6.65 (d, J = 3.1
Hz, 2H), 3.80 (s, 6H), 1.45 (s, 18H). Synthesis of the phosphorochloridite13
A solution of distilled NEt3 (0,929 g, 9,2 mmol) in THF (11,25 mL) was prepared, and PCl3 (0,630 g, 4,9 mmol) was added into this mixture. The prepared diol (1,63 g, 4,5 mmol) was dried with toluene by co-evaporation, dissolved in THF (4,5 mL) and slowly added at -78⁰C to the PCl3 solution. The reaction mixture was stirred for 30 min at -78⁰C and then allowed to warm up to r.t.. The excess of PCl3 was removed in vacuo, by co-evaporation with toluene (3*6 mL). The phosphorchloroidite was obtained as a yellow foam. 1H NMR (300 MHz, CD
2Cl2) δ 7.02 (d, J = 3.1 Hz,
2H), 6.75 (d, J = 3.1 Hz, 2H), 3.83 (s, J = 1.1 Hz, 6H), 1.45 (d, J = 1.1 Hz, 18H). 31P NMR (121 MHz,
CD2Cl2) δ 172.27 (s).
Synthesis of the urea amine12
Ethylenediamine (1,20 g, 20 mmol) was filtrated over neutral aluminium oxide and dissolved in toluene/hexane (1/1, 40 mL). A solution of phenyl isocyanate (1,19 g, 10 mmol) in toluene/hexane (1/1, 20 mL) was prepared and added dropwise to the vigorously stirring solution of ethhylenediamine at 0⁰C. The reaction mixture was allowed to warm up to r.t., the precipitate was filtered and washed two times with n-hexane. Yields of several batches: 66% to 98%. 1H NMR (400 MHz, DMSO) δ 8.12 (s, 1H), 7.47 (d, J =
Synthesis of the phosphoramidite13 (1)
A solution of the phosphorochloridite (1,75 g, 4,13 mmol) in THF (40 mL) was prepared and cooled to 0⁰C. A solution of the urea amine (0,74 g, 4,13 mmol) and NEt3 (0,79 g, 7,89 mmol) in THF (80 mL) were added and stirred for 1 h at 0⁰C. The reaction mixture was allowed to warm uo to r.t. and stirred for 3 h. Followed by filtration of the reaction mixture and evaporation of the filtrate. The solids were dissolved in a small amount of toluene and hexane was added. The formed precipitation was separated from the solvents and washed with hexane resulting in light yellow solids. 1H NMR (300
MHz, CD2Cl2) δ 7.27 (m, 5H), 7.00 (d, J = 3.1 Hz, 2H), 6.70 (d, J = 3.0 Hz, 2H), 6.24 (s, 1H), 4.83 (s, 1H),
3.82 (s, 6H), 3.44 (dt, J = 29.4, 6.6 Hz, 1H), 3.19 (q, J = 5.9 Hz, 2H), 2.87 (m, 2H), 1.49 (s, 18H). 31P NMR
(121 MHz, CDCl3) δ 145.93 (s).
Synthesis of the Pd precursor15
A solution of [(η3-allyl)PdCl]2 (149 mg, 0,4 mmol) and AgBF4 (154 mg, 0,8 mmol) in DCM (4 mL) was prepared and stirred for 15 min. Cyclooctadiene (0,14 g, 1,3 mmol) was added and the mixture was stirred for another 2 min. the reaction mixture was filtered and residue washed two times with DCM (1 mL). Ether (4 mL) was added to the filtrate and the precipitate was filtrated and washed three times with ether (4 mL) and dried on air. The solids were dissolved in DCM (6 mL) and filtered through celite. Solvents were removed in vacuo.
Complex formation
A solution of phosphoramidite 1 (16 mg, 0,030 mmol) and Pd precursor (4,8 mg, 0,014 mmol) in toluene (0,7 mL) was prepared and stirred for 15 min. 31P NMR (121 MHz, CDCl
3) δ 132.32 – 132.15
(m).
Kinetic study: reaction conditions
For the kinetic studies the following conditions were maintained: 0,1 mmol cinnamyl alcohol, 0,15 mmol N-methylaniline, 3 mol% [(η3-allyl)Pd(cod)]BF
4 (palladium precursor), 6 mol% ligand, 6 mol%
urea, 0,5 mL toluene. The kinetic study with ligand 2 started at room temperature and was heated up after 17 hours to 80oC. The other kinetic studies were performed at room temperature.
CD measurements
NMR spectra
9 REFERENCES
1) Muzart, J., Palladium-catalysed reactions of alcohols. Part B: Formation of C–C and C–N bonds from unsaturated alcohols, Tetrahedron, 2005, 61, 4179- 4212
2) Yamamoto, A., Carbon-Oxygen Bond Activation by Transition Metal Complexes, Adv.
Organomet. Chem, 1992, 34, 111-147
3) Trost, B.M., The atom economy--a search for synthetic efficiency, Science, 1991, 254, 1471-1477 4) Rothenberg, G., 'Catalysis: Concepts & Green Applications', Wiley-VCH, Weinheim, 2008,
Chapter 1.1
5) a) Lafrance, M., Roggen, M., Carreir, E. M., Direct, Enantioselective Iridium-Catalyzed Allylic Amination of Racemic Allylic Alcohols, Angewandte, 2012, 51, 3470-3473 b) Das, K., Shibuya,R., Nakahara, Y., Germain, N., Oshima, T., Mashima, K.,Platinum-Catalyzed Direct Amination of Allylic Alcohols with Aqueous Amonia: Selective Synthesis of Primary Allylamines, Angewandte, 2012, 51, 150-154
6) Sundararaju, B., Achard, M., Bruneau, C., Transition metal catalyzed nucleophilic allylic substitution: activation ofallylic alcohols via p-allylic species, Chem. Soc. Rev., 2012, 41, 4467-4483
7) Serveral additives are named in this paper: Banerjee, D., Jagadeesh, R.V, Junge, K., Junge, H., Beller, M., An Efficient and Convenient Palladium Catalyst System for the Synthesis of Amines from Allylic Alcohols, ChemSusChem, 2012, 5, 2039-2044
8) a)Piechaczyk, O., Doux, M., Ricard, L., le Floch, P., Synthesis of 1-Phosphabarrelene Phosphine Sulfide Substituted Palladium(II) Complexes: Application in the Catalyzed Suzuki Cross-Coupling Process and in the Allylation of Secondary Amines, Organometallics 2005, 24, 1204-1213; b)Piechaczyk, O., Thoumazet, C., Jean, Y., le Floch, P., DFT Study on the Palladium-Catalyzed Allylation of Primary Amines by Allylic Alcohol, J. Am. Chem. Soc., 2006, 128, 14306-14317; c)Thoumazet, C., Grützmacher, H., Deschamps, B., Ricard, L., le Floch, P., Testing Phosphanes in the Palladium Catalysed Allylation of Secondary and Primary Amines, Eur. J. Inorg. Chem., 2006,
19, 3911- 3922; d)Mora, G., Deschamps, B., van Zutphen, S., Le Goff, X. F., Ricard, L., Le Floch, P.,
Xanthene-Phosphole Ligands: Synthesis, Coordination Chemistry, and Activity in the Palladium-Catalyzed Amine Allylation, Organometallics, 2007, 26, 1846-1855
9) a) Usui, I., Schmidt, S., Breit, B., Dual Palladium- and Proline-Catalyzed Allylic Alkylation of Enolizable Ketones and Aldehydes with Allylic Alcohols, Org. Lett., 2009, 11, 1453-1456; b) Usui, I.,Schmidt, S., Keller, M., Breit, B., Allylation of N-Heterocycles with Allylic Alcohols Employing Self-Assembling Palladium Phosphane Catalysts, Org. Lett., 2008, 10, 1207-1210
10) Spessard, G. O., Miesssler, G.L., Organometallic Chemistry, Prentice Hall, 1997, Chapter 5.1.2
11) Kelly, S.M., Price, N.C., The Use of Circular Dichroism in the Investigation of Protein Structure and Function, Current Protein and Peptide Science, 2000, 1, 349-384
12) Jana, R., Tunge, J.A., A Homogeneous, Recyclable Polymer Support for Rh(I)-Catalyzed C-C Bond Formation, J. Org. Chem., 2011, 76, 8376–8385
13) Breuil, P-.A.R., Monodentate, Supramolecular and Dynamic Phosphoramidite Ligands Based on Amino Acids in Asymmetric Hydrogenation Reactions, Ph.D., 2009, University of Amsterdam, The Netherlands; Method A in Chapter 2 is modified to these substrates
14) Meeuwissen, J., Urea-functionalized Phosphorus Ligands in Transition Metal Catalysis, Ph.D., 2009, University of Amsterdam, The Netherlands, General procedure of urea-amines, p. 41. 15) White, D.A., Cationic complexes of cycloocta-1,5-diene with palladium(II) and platinum(II),
