Chapter 5 124

Application of LC-NMR for the quick identification of process

related impurities in the synthesis of Docetaxel intermediate

5.1 Introduction

Taxol, a naturally occurring diterpene, has showed

overwhelming evidence of the chemotherapeutic value [1 & 2] in the

clinical trials against cancer by the year 1988. Consequently, a

variety of synthetic analogues of taxol were synthesized and examined

for their ability to inhibit microtubules disassembly and their

potencies were compared [3[. Some of these compounds showed a

good correlation between cytotoxicity and the microtubule

disassembly assay [4].

In taxoid family Paclitaxel (i.e. Taxol) which was initially named

as Taxol and Docetaxel (i.e. Taxotere) (Scheme 5.1) are currently

considered to be two of the most exciting drugs in the cancer

chemotherapy. These two molecules exhibit significant antitumor

activity [5].

Paclitaxel (i.e. Taxol) is used for the treatment of breast, ovarian,

lung, bladder, prostate, melanoma, esophageal, as well as other types

of solid tumor cancers. Taxotere can be used for lonely or with

combination mode for the following cancers Breast Cancer, Advanced

Non-Small Cell Lung Cancer, Metastatic Androgen-Independent

Prostate Cancer Advanced Gastric/GE Junction,Locally Advanced

Head and Neck Cancer

Chapter 5 125

Scheme 5.1 Structures of Paclitaxel and Docetaxel

5.2 Present Work:

A number of taxol analogs have been described by F. Guueritte-

Voegelein et al [6] Docetaxel (DCT) is an antineoplastic agent. The

Docetaxel was synthesized and marketed by Dr. Reddy’s Laboratories

Ltd, for the treatment of cancer. The total synthesis of Docetaxel, is

very lengthy and difficult to manufacture. There are several reports

available for the synthesis Docetaxol from naturally occuring 10-

deacetylbaccatin III (10-DAB III), which is readily available from the

renewable leaves of Taxus baccata (European Yew) and it was

described in Scheme 5.2 [7].

Chapter 5 126

Scheme 5.2 Synthetic scheme of Docetaxel

For the preparation of Docetaxol, the 10-DAB III hydroxyl

groups at 7 and 10 positons have to be protected. Due to the steric

hinderence the hydroxyl group at 1 position will not react with the

protecting groups. However the present reaction conditions are not

favorable for the substitution at 13 position. Based on the literature

precedence N, N-carbodiimidazole was chosen as protecting group.

During the process development / optimization of Docetaxel, few by-

products were observed in the preparation of DCT-1. The synthesis of

DCT-1 was given in Scheme 5.3. These by-products were observed at

higher percentages when acetone was used as a solvent. To know the

structure of these by-products LC-MS and LC-NMR studies were used

without isolation. The characterization of these by-products in the

synthesis of docetaxel process development is presented in this

chapter.

Chapter 5 127

Scheme 5.3 Preparation of DCT-1

5.3 Experimental

5.3.1 Samples

Crude sample of DCT-1 was subjected for the identification of

the by-products. Samples of pure 10-DAB III and DCT-1 were used to

generate spectral data for the comparison with those of the by-

products. Pure samples of 10-DAB III and DCT-1 were obtained from

Dr. Reddy’s Laboratories Ltd., Hyderabad, India.

5.3.2 Analytical HPLC

An in-house HPLC method was developed for the separation of

the impurities and the products. The HPLC conditions are shown in

Table 5.1.

Chapter 5 128

Table 5.1 Analytical HPLC conditionsColumn 250mm× 4.6mm, 5 μm, RPB

(Highchrom , manufactured by Hichrom Ltd.,Berkshire, UK)

Solvent – A Water (HPLC grade)

Solvent – B Acetonitrile (HPLC grade)

Gradient Program(T / %B)

0/35, 15/65, 25/75, 30/95, 35/100, 39/100 and40/35 for post run

Flow rate 1.0 mL / min

Detection 230 nm

Injection volume 10 μl

Diluent 1:1 (water : CH3CN)

5.3.3 Mass spectrometry

A HPLC method suitable for LC-MS was developed by using

Ammonium Acetate buffer. The LC conditions are shown in Table 5.2.

Table 5.2 LC-MS conditions

Column Inertsil ODS 4.6mm 250 mm, 5.0 m

Solvent – A 0.01M Ammonium Acetate (pH 3.2 with AceticAcid)

Solvent – B Acetonitrile (HPLC grade)

Isocratic 50:50

Flow rate 1.0 mL / min

Detection 230 nm

Injection volume 10 μl

Diluents 1 : 1 (water : CH3CN)

Chapter 5 129

5.3.3 LC-NMR spectroscopy

The regular HPLC method (cf. Table 5.1) was modified for the LC-

NMR analysis to reduce the retention time. The LC conditions are

shown in Table 5.3.

Table 5.3 LC conditions for LC-NMR

Column Hypersil BDS C18, 205mm 4.6 mm, 5.0 m(Thermo Hypersil-Keystone, Germany)

Solvent – A 0.01M Potassium dihydrogen phosphate(pH 3.2 with H3P=O4) in D2O

Solvent – B Acetonitrile (LC-NMR grade)

Isocratic 50:50

Flow rate 1.0 mL / min

Detection 230 nm

Injection volume 10 μL

Diluents 1 : 1 (water : CH3CN)

NMR Spectroscopy : The LC - NMR spectra were recorded in at

300K using Varian Unity INOVA 500 MHz NMR spectrometer,

equipped with a Varian 3mm Interchangeable Flow Cell (IFC) 1H-

19F{13C/15N} pulse field gradient (PFG) probe fitted with a 60l flow

cell. The 1H chemical shift values were reported on the scale in

ppm, relative to acetonitrile ( = 1.95). Standard ‘lc1d’ pulse sequence

provided by Varian was used, 500 MHz 1H-19F{13C/15N} 60l PFG

microflow Probe. The LC – NMR was run by using stop flow method.

Chapter 5 130

5.3.4 NMR spectroscopy

The NMR experiments, for the pure 10-DAB III and DCT-1, such

as 1H, 13C, gDQCOSY, gHSQC, gHMBC and NOESY were performed on

Mercury plus 400 MHz and Unity Inova 500 MHz, Varian instruments,

at 25 C in DMSO-d6.

5.4 Results and discussions

5.4.1 Detection and Separation of by-products.

The analytical HPLC chromatogram of the crude DCT-1 was

obtained by using the HPLC method (cf. Section 5.2.2) is shown in

Figure 5.1. Two unknown by-products were observed at 7.81 and

9.35 min. Henceforth, these by-products will be referred as BP-I and

BP-II. The starting compounds, (10-DAB III) and the product (DCT-1)

eluted at 5.68 and 11.10 min respectively.

Figure 5.1 Analytical HPLC chromatogram of crude DCT-1

Chapter 5 131

Figure 5.2 LC-MS chromatogram of crude DCT-1(a) TIC spectrum (b) UV Chromatogram

The LC-MS data was generated (Figure 5.2) to obtain the

molecular ion information of BP-I and BP-II. Interestingly, both BP-I

and BP-II displayed the same molecular ion, at m/z 639 (Figures 5.3

and 5.4) [M+H]+, corresponding to mono protected 10-DAB III. While

the product DCT-1 has both 7 and 10-hydroxyl groups are protected.

The proposed structures of BP-I and BP-II are shown in Scheme 5.4.

LC-NMR technique is employed for the unambiguous identification of

by-products, BP-I and BP-II.

Chapter 5 132

Scheme 5.4 Proposed structures of by-products

5.4.2 Structural chemistry of 10-DAB II, DCT-1, BP-I and BP-II

As a first step, the NMR structural investigation of DCT-1 is

taken up. This would facilitate easier identificaion of the structures of

by-products. The positive ES-MS spectrum (Figure 5.5) of DCT-1 gave

molecular ion at 733.3, confirms substitution of two imidazolyl

moieties on 10-DAB III. The proton NMR assignments of 10-DAB III

were already reported [8]. The 1H, 13C, DQCOSY and HMBC NMR data

of DCT-1 generated in DMSO-d6. The spectra are shown in Figures 5.6

to 5.9. The 1H and 13C NMR data of 10-DAB III, was also generated in

DMSO-d6 for comparison with DCT-1 (Figures 5.10 and 5.11).

Chapter 5 133

Figure 5.3 Positive ES-MS spectrum of 5.36min

Figure 5.4 Positive ES-MS spectrum of 6.69min

Chapter 5 134

Figure 5.5 Positive ES-MS spectrum of DCT-1

Figure 5.6 1H NMR spectrum of DCT-1 in DMSO-d6.

Chapter 5 135

Figure 5.7 13C NMR spectrum of DCT-1 in DMSO-d6.

Figure 5.8 COSY spectrum of DCT-1 in DMSO-d6.

Chapter 5 136

Figure 5.9 HMBC spectrum of DCT-1 in DMSO-d6.

The numbering scheme for 10-DAB III and DCT-1 used in the

discussion is shown in Scheme 5.5.

Scheme 5.5 Structures of 10-DAB III and DCT-1 with numbering

The NMR assignments (Table 5.4) of 10-DAB III and DCT-1 were

compared. The substitution of imidazolyl groups, at C-7 and C-10,

lead to changes in the chemical shifts of protons and carbons in DCT-

1. The chemical shifts of methine proton (carbon) at C-7 position in

10-DAB-III, 4.11 ppm (70.94 ppm) have shifted to 5.62 ppm (75.85

Chapter 5 137

ppm) in DCT-1. Similarly, the proton (carbon) chemical shifts at C-10

positions in 10-DAB-III, 5.13 ppm (74.39 ppm), have shifted to 6.39

ppm (78.72 ppm) in DCT-1. These chemical shift perturbations due to

substitution can be clearly seen in the comparison of 1H and 13C NMR

spectra (Figures 5.10 and 5.11). It is clear that these changes in

chemical shifts of the protons (and carbon) at C7 and C-10 can be

used as a diagnostic tool for the identification of the site of the

imidazolyl substitution.

Figure 5.101H NMR spectra overlay of 10-DAB III & DCT-1

Chapter 5 138

Figure 5.11 13C NMR spectra overlay 10-DAB III (top) & DCT-1(bottom)

Table 5.4 NMR assignments of 10-DAB III and DCT-1

10-DAB DCT-1

No1 1H 1H (ppm) 13C(ppm) 1H (ppm) 13C(ppm)

1 - - 76.97 - 76.861 OH 4.32 4.752 1H 5.41 74.88 5.51 74.02

3 1H 3.80 46.54 3.90 46.38

4 - - 80.11 -- 79.40

5 1H 4.92 83.77 5.06 82.68

6 Ha 2.31 36.57 2.73 32.62Hb 1.65 1.96

7 1H 4.11 70.94 5.62 75.857 OH 4.96 - -8 - - 57.07 -- 55.669 - - 210.35 -- 202.0610 1H 5.13 74.39 6.39 78.7710 OH 4.73 -11 - 134.48 - 128.8112 - - 141.61 - 149.14

Chapter 5 139

13 1H 4.63 66.09 4.72 66.0813 OH 5.19 5.55 --14 2H 2.16 40.00 2.30 39.66

2.2415 - - 42.46 -- 42.3316 3H 0.94 26.76 1.07 26.4017 3H 0.95 20.16 1.03 20.8418 3H 1.90 14.78 2.04 15.1019 3H 1.53 9.70 1.82 10.5320 Ha 4.05 75.48 4.15 75.27

Hb 4.03 4.1221 - 169.53 -- 170.2722 3H 2.20 22.28 2.27 22.1923 - - 165.26 -- 165.2424 - - 130.33 -- 129.95

25,29 2H 8.02 129.51 8.05 129.5926,28 2H 7.56 128.62 7.60 128.72

27 1H 7.66 133.14 7.70 133.3830 - - - -- 147.1531 1H - - 7.09 130.6732 1H - - 7.30 117.1933 1H - - 7.99 135.1034 - - - -- 147.3835 1H - - 7.06 129.9236 1H - - 7.45 117.6037 1H - - 8.08 129.95

The 1H NMR assignments of 10-DAB III and DCT-1 (both in

DMSO-d6) and LC-NMR along with the BP-I and BP-II were shown in

Table 5.5, between 4.50 – 8.50 ppm.

Chapter 5 140

The overlaid LC-NMR spectra of DCT-1 and 10-DAB-III are

shown in Figure 5.12. The chemical shift changes due to substitution

discussed earlier are indicated by arrows in the Figure 5.12.

Having the NMR assignments of 10-DAB III and DCT-1 as

basis, the LC-1H NMR data of arising from LC peaks due to BP-I and

BP-II has been analyzed to arrive at their structures. The LC 1H NMR

data of BP-I and BP-II was compared with those of the 10-DAB III and

DCT-1.

Table 5.5 1H NMR Assignments of 10-DAB III, DCT-1, BP-I and BP-IIfrom LC-NMR

10-DAB III DCT-1 BP-I BP-II

No. 1H (ppm) (ppm) (ppm) (ppm)2 1H 5.45 5.55 5.50 5.653 1H 3.80 3.90 3.75 3.955 1H 4.95 5.05 4.95 5.007 1H - 5.55 - 5.4410 1H 5.20 6.38 6.50 5.3013 1H 4.75 4.82 4.80 4.75

25,29 2H 8.02 8.05 8.05 8.0526,28 2H 7.50 7.55 7.55 7.50

27 1H 7.64 7.65 7.65 7.6531 1H - 7.00 7.05 -32 1H - 7.30 7.50 -33 1H - 7.99 8.25 -35 1H - 7.00 - 7.0036 1H - 7.40 - 7.4037 1H - 8.10 - 8.15

Chapter 5 141

Figure 5.12 Overlay of LC-NMR spectra of DCT-I with 10-DAB III

5.4.3 Structure confirmation of BP-I

The 1H NMR spectra of DCT-1, 10-DAB III and BP-I obtained by

LC-NMR are compared in Figure 5.13. A baseline drift is observed in

the range of 1.5 – 4.5 ppm. This is due to the presaturation of the

signals due to acetonitrile and water at 1.9 and 4.2 ppm respectively.

The aromatic region of BP-I has showed signals at 7.05 (1H, s),

7.50 (1H, s) and 8.25 ppm(1H, s) in addition to the signals due to

phenyl moiety (Figure 5.13). This observation showed that there is

only one imidazolyl moiety in the BP-I which is in conformity with the

molecular weight information from mass spectrometry.

Chapter 5 142

Figure 5.13. Overlay-LC-NMR spectra of DCT-1, BP-I and 10-DAB III

The following significant changes can be observed in the 1H

chemical shifts of BP-I. The singlet at 5.20 ppm which is due to

methine proton at C-10 in the 10-DAB III spectrum has shifted to 6.50

ppm in BP-I. As discussed earlier, this is diagnostic of substitution of

imidazolyl group at C-10. Interestingly, the chemical shift of methine

proton at C-10 in DCT-1 is 6.39 ppm which is quite comparable to the

chemical shift in BP-I. No other significant changes were observed in

the spectrum of BP-I. All these observations confirm the structure of

BP-I to be 10-imidazolyl substituted 10-DAB III.

Chapter 5 143

Scheme 5.6 Structures of 10-DAB III and BP-I

5.4.4 Structure confirmation of BP-II

The 1H NMR spectra of DCT-1, 10-DAB-III and BP-II obtained by

LC-NMR are compared in Figure 5.14.

Figure 5.14.Overlay-LC-NMR spectra of DCT-1, BP-II and 10-DAB III

Chapter 5 144

The aromatic region of BP-II has showed signals at 7.00 (1H, s),

7.40 (1H, s) and 8.15 (1H, s) in addition to the signals due to phenyl

moiety (Figure 5.13). This observation showed that there is only one

imidazolyl moiety in BP-II which is in conformity with the molecular

weight information from mass spectrometry.

The following significant changes can be observed in the 1H

chemical shifts of BP-II. The proton NMR spectrum of aliphatic region

of BP-II showed a signal at 5.44 ppm (doublet of a doublet) integrating

for one proton. This could be assigned due to the methine proton

connected to C-7. The corresponding methine proton in 10-DAB III

which is expected at ~4.1 ppm could not be seen as it is under the

signal due to water, and hence cannot be seen. The C-7 methine

proton signal was observed at 4.11pm in 1H NMR spectrum collected

in DMSO-d6. The remaining protons didn’t show significant chemical

shift changes. Interestingly, the chemical shift of methine proton at

C-7 in DCT-1 is 5.55 which is quite comparable to the chemical shift

in BP-II.

All these observations confirm the structure of BP-II to be 7-

imidazolyl substituted 10-DAB III.

Chapter 5 145

Scheme – 5.7 Structure of 10-DAB III and BP-II

5.5 Formation of by-products

The synthesis of DCT-1 was to protect 7 and 10 positions of 10-

DAB III with N, N-carbodiimidazole. The reaction is shown in Scheme

5.3.

The elucidated structures of BP-I and BP-II suggest that these

products have formed due to partial protection at either 7 or 10 in 10-

DAB III (Scheme 5.8)

O

H3C

OH

O

CH3CH3

CH3

O

OH 3C

O

O

HO

OH

HO H

H3CO

O

O

HO

HO

OO

H3C

O

O

O

N

O

N

N

O

N

H3CO

OH

O

HO

HO

OO

H3C

O

O

O

N

O

N

H3CO

O

OH

HO

HO

OO

H3C

O

O

O

N

O

N

N,N-Carbonyl dimidazoleEthylAceta te RT 3.5hr

10 DAB

DCT-1

+ +

Scheme 5.8 Formation of by-products BP-I and BP-II

Chapter 5 146

5.6 Conclusion

The LC-MS data has helped in the identification of the tentative

structures of the by-products formed during the reaction. The LC-

NMR has helped in the quick identification of the structures of these

by-products unambiguously without resorting to isolation of the by

Preparative HPLC.

Chapter 5 147

5.7 REFERENCES

1 Rowinsky, E.K., Cazenave, L., and Donehower, R., J. Natl.

Cancer Inst., 82, 1990, 1247-1259.

2 Guéritte - Voegelein,F., Guénard, D., Lavelle, F., Le Goff, M.,

Mangatal, L., and Potier, P., J. Med. Chem., 34, 1991, 992-998.

3 Rowinsky, E.K. and Donehower, R.C., J. Natl. Cancer Inst., 83,

1991, 1778-1781.

4 Guénard, D., Guéritte-Voegelein,F., Dubois, J., and Potier,P.,

Structure activity relationships of taxol and taxotere analogues,

presented at the 2nd Natl. Cancer Institute Workshop on Taxol

and Taxus, Alexandria, VA, September 1992.

5 E. Tomiak, M.J. Piccart, J. Kerger, D. Devaleriola, E. Tueni, D.

Lossignol, S. Lips and M. Bayssas, Eur. J. Cancer, 27 (suppl.2),

1991, 1184.

6 F. Gue- ritte-Voegelein, D. Guenard, F. Lavelle, M. T. LeGoff, L.

Mangatal, P. Potier, J. Med. Chem., 34, 1991, 992-998.

7 F. Lavelle, C. Fizames, F. Guéritte-Voegelein, D. Guénard and P.

Potier, Proc. Am. Ass. Cancer Res., 30, 1989, 566.

8 David G. I. Kingston, Douglas R. Hawkins, and Liza Ovington, J.

Nat. Products, 45, 4, 1982, 466.