Effect oF β-Alanine on The Preparation of 4-Ethoxy-Cinnamic Acid

It is well-known that Verley-Doebner modification is a high efficient method for the preparation of cinnamic acid derivaties, especially with the compounds containing electron-donating subtituents at para position in aromatic ring. In this paper, 4-ethoxycinnamic acid was synthesized according to Verley-Doebner reaction with the use of pyrine acting as catalyst and solvent. The effect of the β-alanine concentration on the conversion of the starting material of 4-ethoxy-benzaldehyde was thoroughly investigated using high performance liquid chromatography. The results showed that consuming of 8% of β-alanine would convert 100% of 4-ethoxybenzaldehyde to 4-ethoxy-cinnamic acid.


INTRODUCTION
Cinnamic acid is an anti-bacteria and anti-mould organic compound which has been widely used in cosmetic and food industries.In particular, its derivatives are also the vital source of pharmaceutical products which can be applied in the high pressure treatment and tumor inhibition [1][2][3][4].Beside the application in pharmaceutical field, cinnanic acid and its derivatives have been also used as "building blocks" for the synthesis of photo-sensitive polymers due to the corrosion inhibitors [5 -15].
The preparation of cinnamic acid was pioneered by the work of Perkin' group which is well-known as Perkin reaction [16].In this reaction, benzaldehyde reacted with malonic anhydride in the presence of sodium acetate or potasium acetate as catalyst.In order to obtain a rather good conversion of benzadehyde, long reaction time (8 h) and high temperature (180 °C) should be applied [16].This reaction is unsuitable for the synthesis of cinnamic acid derivatives with electron-donating substituents in the aromatic ring.For example, the yields of 4-methoxy-cinnamic acid and 4-ethoxy-cinnamic acid were 30% and 36%, respectively.In order to obtain relatively good yields of cinnamic acid derivaties with electron-donating subtituents at the para position in the aromatic ring, many literatures have focused on the Knoevenagel modification using malonate ester, malonic acid or acetic acid insteads of malonic anhydride [17][18][19][20][21][22][23][24].Different types of catalysts including acidic and basic catalysts were used in this approach, which showed a prospective potential for the synthesis of many types of cinnamic acid derivaties, especially compounds with the electron-donating subtituents at para position in the aromatic ring.For intance, when carrying out the reation between 4-methoxy-benzaldehyde with acetic acid in the solvent of N-methyl-2-pyrrolidone using boron-based catalysts (NaBH 4 , BBr 3 , or LiB 4 O 7 ) [17][18][19] at 180-190 °C for 12 hours, the yield of 4-methoxy-cinnamic acid was in the range of 53-64%.In other literatures, microwave was applied instead of tradditional heating, which reduced the reaction time to several minutes and yielded high percentage of 4-methoxy-cinnamic acid [20][21][22][23][24]. Recently, Verley-Doebner modification with the use of malonic acid and β-alanine as co-catalyst in the presence of pyridine showed a great potential for the preparation of a wide range of cinnamic acid derivatives [25][26][27][28][29].In this article, we also used Verley-Doebner modification for the preparation of 4-ethoxy-cinnamic acid with thoroughly investigation of the effect of βalanine concentration on the conversion of 4-ethoxy-benzaldehyde which has not yet been mentioned in the literature.In addition, with the proposed experiment procedure, the yield of 4-ethoxy-cinnamic acid was high (>98.2%)with the short reaction time.

Preparation of 4-Ethoxy-cinnamic Acid
The experiment was set up followed the Scheme.(1) in which all of the reactions were carried out under nitrogen atmosphere.A typical reaction (Scheme.2), Run 1 of Table 1 was described as follows.To a 100 mL of two neck round-bottom flask were added 1853 mg of malonic acid (17.8 mmol) and 51.3 mg of β-alanine (0.58 mmol).The solid mixture was then kept for overnight under reduced pressure using vacuum pump.The vacuum atmosphere was then switched to nitrogen atmosphere thoroughly with at least three cycles of vacuuming and purging nitrogen.To the above solid mixture were added 1.00 mL of 4-ethoxy-benzaldehyde (7.20 mmol) and 3.20 mL of pyridine (39.7 mmol) under nitrogen atmosphere.The reaction mixture was then heated to the reflux condition (120 °C) using oil bath and hotplate.During reaction period, samples were taken for high performance liquid chromatography measurements at 0, 1, 2, 3, 4, 5, 7, 10, 15, and 30 mins for investigating conversion of 4-ethoxy-benzaldehyde.Reaction was stopped at 30 min and the reaction mixture was then cooled to room temperature.5 mL of concentrated HCl was then slowly added to the mixture which was placed in ice water.The precipitated white crystals were filtered, washed with cold water and dried under vacuum for overnight (98.2% yield).For analysis, white crystals were recrystallized twice using ethanol and water.

Investigation on the Conversion of 4-Ethoxy-Benzaldehyde Using High Performance Liquid Chromatography
The conversion of 4-ethoxy-benzaldehyde was investigated using High Performance Liquid Chromatography (HPLC) measurements at different reaction times of 0, 1, 2, 3, 4, 5, 7, 10, 15 and 30 mins. 10 μL of samples at a certain time were diluted with 1 mL of ethanol.The mixtures were then applied for HPLC analysis with the mobile phase of the acetonitrile/water mixture (60/40 v/v), flow rate of 1 mL/min, C18 column, 1 μL of injected volume, and UV detector at 254 nm.The peaks appear at 2.39 min in the spectra of HPLC corresponding to 4-ethoxy-benzadehyde.The conversion of 4-ethoxy-benzadehyde was calculated using the following equation: Where Conv ethoxy : Conversion of 4-ethoxy-benzadehyde; A 2.39 and A t 2.39 : peak areas of 4-ethoxy-benzadehyde at 0 and t min, respectively.

Measurements
Fourier Transform Infrared Spectroscopy (FT-IR) was measured under KBr pellets using Bruker's VERTEX 70 at a resolution of 2 cm -1 .Raman measurements were carried out using Horiba XploRA laser 785 nm.Gas chromatography-Mass spectroscopy (GC-MS) was carried out using Agilent 7890B for GC and Agilent 5977A for MS with the mobile phase at 1 mL/min of flow rate.

Effect of β-Alanine Concentration on the Conversion of 4-Ethoxy-Benzaldehyde
In this paper, the conversion of 4-ethoxy-benzadehyde was thoroughly investigated using HPLC measurements.

A A Conv t ethoxy
HPLC profiles at different reaction conditions with different amounts of β-alanine were described in Fig. (1a, b, c).The peaks at 1.87 min and 2.39 min correspond to the desired product of 4-ethoxy-cinnamic acid and the reactant of 4ethoxy-benzadehyde, respectively.In the case of using 2% of β-alanine Fig. (1a), the reactant remained at high level in the reaction mixture even at long reaction time, indicating the used amount of β-alanine was insufficient.The amount of β-alanine was therefore increased to 4% leading to a better result shown in Fig. (1b); however, 4-ethoxy-benzadehyde still remained after 30 min of reaction time.In the case of using 8% of β-alanine Fig. (1c), the 4-ethoxy-benzadehyde was completely consumed after 5 min of reaction time.The results showed in Fig. (1d) obviously demonstrated that the amount of β-alanine affected the conversion of 4-ethoxy-benzadehyde.In particular, when using 2% of β-alanine, the conversion of 4-ethoxy-benzadehyde was below 45% and required 30 min to reach maximum value of 42.8%.When increasing β-alanine concentration to 4%, the conversion of 4-ethoxy-benzadehyde went up to 87.9% and required less time (15 mins) to reach maximum value as compared with the case of using 2% β-alanine.In the case of using 8% βalanine, the conversion of 4-ethoxy-benzadehyde reached 100% rapidly after 5 min of reaction time.

Characterization of 4-Ethoxy-Cinnamic Acid
The obtained product of 4-ethoxy-cinnamic acid was recrystallized twice using ethanol and water for characterization.The FT-IR and Raman spectra of 4-ethoxy-cinnamic acid were showed in Fig.  2b) and the vibration assignments were described in Table 2 followed the literatures [30,31].

Vibrations of Carboxylic Acid
The broad band in the range of 3200-2600 cm -1 in FT-IR spectrum is characteristic for the vibration of O-H stretching in carboxylic acid.Other strong peak appears at 1676 cm -1 was assigned for the C=O stretching mode.This peak is strong in FT-IR spectrum but completely absented in Raman spectrum.The peak at 1335 cm -1 is assigned for C-O stretching.The C-COOH stretching vibration was confirmed with the peak appearing very strong at 1308 cm -1 in the FT-IR spectrum.The OH in-plane bending vibration appears with strong intensity at 1117 cm -1 in the FT-IR spectrum, while OH out-of-plane bending shifted to lower wavenumber at 547 cm -1 .All of the peaks corresponding to the vibrations of carboxylic acid are marked by red asterisks in (Fig. 2).
The Open Materials Science Journal, 2018, Volume 12 63

Vibrations of -C=C-and C-C
The vibration of C-C stretching in phenyl ring is obviously observed both in FT-IR and Raman spectra with the intensity ranging from weak to very strong which are marked by closed blue circles in the Fig. (2).These peaks appear in the region of 1650-1100 cm -1 .In FT-IR spectrum they can be seen at 1600, 1573, 1510, and 1475 cm -1 , whereas at 1597, 1575 cm -1 in Raman spectrum.Other characteristic vibration of C-C bond in phenyl ring is CCC in-planebending, which appears in FT-IR spectrum at 785, 582, and 515 cm -1 .For C=C stretching in double bond, the characteristic peaks appear very strong in both FT-IR and Raman spectra; specifically at 1622 cm -1 in FT-IR spectrum and at 1626 cm -1 in Raman spectrum, which are marked by open blue circles in the (Fig. 2a).

Vibrations of C-H
The characteristic vibrations of C-H in phenyl and double bond are stretching, in-plane bending and out-of-plane bending, which are marked by open green diamonds in the Fig. (2).The stretching mode shows weak peak in FT-IR spectrum at 3037 cm -1 .The in-plane C-H bending mode in phenyl ring appears at 942 cm -1 in FT-IR spectrum.The outof-plane C-H bending vibrations in phenyl ring appear at 858, 835, 802, 743, 692, and 673 cm -1 in FT-IR spectrum and at 861, 749, 672 cm -1 in Raman spectrum.The peak observed at 835 cm -1 is characteristic peak for the di-substituent at para position of phenyl ring.For the vibrations of C-H in double bond, in-plane bending and out-of-plane bending modes were observed.Specifically, the in-plane bending mode of C-H in double bond appears very strong at 1285 and 1213 cm -1 in FT-IR spectrum and medium at 1246 and 1208 cm -1 in Raman spectrum.Whereas, the out-of-plane bending of C-H in double bond appeared at 980 cm -1 in FT-IR spectrum and 981 cm -1 in Raman spectrum.These are characteristic peaks for the trans isomer of 4-ethoxy-cinnamic acid.

Vibration of CH 3 and CH 2
The vibrations modes of CH 3 and CH 2 in 4-ethoxy-cinnamic acid are CH 3 and CH 2 deformation, CH 2 rocking, and CH 3 wagging which are marked by closed green diamond in the Fig. (2).The CH 2 deformation appeared strongly at 1427 cm -1 in FT-IR spectrum, whereas CH 3 deformation appeared at 1396 cm -1 in FT-IR spectrum.The CH 2 rocking vibration is confirmed with the medium peak at 1157 cm -1 in FT-IR spectrum.The CH 3 wagging mode is the characteristic vibration of CH 3 and appears strongly both in FT-IR and Raman spectra at 1174 cm -1 and 1169 cm -1 , respectively.

Vibrations of Ather
The structures of 4-ethoxy-cinnamic acid consist of ether linkages between ethoxy and phenyl, which lead to two stretching vibration modes of νC-OCH 2 and νO-CH 2 which are marked by open purple triangles in the Fig. (2).νC-OCH 2 appears in FT-IR spectrum with high intensity at 1248 cm -1 , while in Raman spectrum at 1246 cm -1 .These vibrations are also present in the Raman spectrum at 1246 cm -1 .In contrast, the νO-CH 2 vibration is observed in the FT-IR and Raman spectra at 1045 cm -1 and 1049 cm -1 , respectively.

GC-MS Measurements
The structure of 4-ethoxy-cinnamic acid was further confirmed using GC-MS measurement and the result is shown in Fig. (3).According to literature [32], the fragmentation of cinnamic acid in the ion source should lead to positively charged ions; hence m/z values in Fig.   3) is attributed to the contribution of natural abundance of 2 H, 13 C or 17 O isotopes to chemical molecules of 4-ethoxy-cinnamic acid.The isotope ratios of (M+1) + and M + of 4-ethoxy-cinnamic acid was theoretically calculated using the below equation: ( The theoretical isotope ratio of (M+1) + and M + for 4-ethoxy-cinnamic acid calculated from equation ( 2) is 12.18%.The observed isotope ratio of 4-ethoxy-cinnamic acid was calculated based on the intensity of the peak at m/z = 193 divided by that of the peak at m/z = 192 in Fig. (3).The results show that the calculated ratio was completely matched to the observed one for 4-ethoxy-cinnamic acid.Therefore, the chemical composition of 4-ethoxy-cinnamic acid (C 11 H 12 O 3 ) was confirmed.The fragmentation of 4-ethoxy-cinnamic acid was initiated from the "mother" ions as showed in Scheme.(3).According to the previous literature [30], the fragmentation route should be classified via three competitive pathways: i) cleavage of OH from carboxylic acid group (route A); ii) elimination of H in phenyl ring and ring-closing (route B) and iii) elimination of C 2 H 4 of ethoxy group forming 4-hydroxy-cinnamic acid ions (route C).For route A, a ion at the m/z = 175 appeared with very low intensity.The further fragmentation of this ion to b ion should result in the appearance of the peaks at m/z = 147 in Scheme.(3).Here, the m/z at 147 in Scheme.(3) should be investigated whether it can be assigned for b ion.The isotope ratio (M+1) + and M + of b ion was calculated theoretically based on equation ( 2) and compared with one obtained from the ratio of the observed intensity of the peaks at m/z = 148 and 147 in Fig. (3).The result shows that the calculated isotope ratio based on b ion was different from the observed ratio.Therefore, the peak at m/z = 147 should not be assigned for b ion.Actually, in route C, there is the f ion with m/z = 147 and its isotope ratio was fitted with the observed ratio in Fig.  3) in the m/z region of 75-164 was similar to that of 4-hydroxy-cinnamic acid which was reported in the literature [32].
(3).Hence, the peak at m/z = 147 in Scheme.(3) should be assigned for f ion in route C.For route B, the peaks at 191 corresponding to c ion and their "children" ions of d (m/z = 163) should appear in Fig.(3); however, its observed isotope ratio was not assigned for d ion.The fragmentation via Route C undergoes the elimination of C 2 H 4 from the structures of 4-ethoxy-cinnamic acid leading to the formation of 4-hydroxy-cinnamic acid ion.This route was strongly confirmed by the appearance of the peak at m/z = 147 in Fig.(3) with strong intensity, which should be assigned for f ion.Interestingly, the theoretical isotope ratio of f ion was in agreement with the observed ratio calculated from the intensity of the peaks at m/z = 148 and 147 in Fig. (3).All other "children" ions belonging Route C fragmentation appear in Fig. (3) at m/z = 147, 135, 119, 107, 91, and 77.Moreover, the GC-MS profile of 4-ethoxy-cinnamic acid in Fig. (