Temperature-resolved thermal analysis of cisplatin by means of Li+

ion

attachment mass spectrometry

Seiji Takahashi,a Yuki Kitahara,a Megumi Nakamura,b Yoshiro Shiokawab and

Toshihiro Fujii*a

Received 9th November 2009, Accepted 2nd February 2010

First published as an Advance Article on the web 24th February 2010

DOI: 10.1039/b923454c

Li+ ion attachment mass spectrometry (IAMS) was evaluated as an analytical methodology for

measurement of the thermally labile, nonvolatile, and insoluble compound cisplatin, which is used

as an anticancer agent in the treatment of testicular and ovarian cancers. We aimed to develop an

improved method for the mass spectrometric determination of cisplatin, particularly in its

molecular ion form. A uniquely designed quadrupole mass spectrometry system along with a Li+

ion attachment technique and a direct inlet probe provided cisplatin molecular ions as Li+ ion

adducts; to our knowledge this is the first reported instance of cisplatin Li+ ion adducts.

Full-scan spectra were obtained with B10 mg samples. Infrared image furnace-ion attachment

mass spectrometry (IIF-IAMS) also was used to study the temperature-programmed

decomposition of this drug. The slope of the plot of signal intensity versus temperature for

cisplatin decomposition from 225 to 249 1C was used to determine an apparent activation energy

(Ea) of 38.0 kcal mol�1 for the decomposition of cisplatin. This decomposition parameter is useful

for predicting drug stability (shelf life). In this study, we have demonstrated that IAMS can be a

valuable technique for the direct mass spectral analysis and kinetic study of d-metal complex

platinum anticancer agents.

Introduction

Cisplatin (cis-diamminedichloroplatinum, Pt(NH3)2Cl2) and

its less toxic analogue are known to be exceptionally useful

anticancer agents. Presently, cisplatin is the drug of choice in

the treatment of various solid tumors, including testicular and

ovarian cancer tumors. Accordingly, the interest in accurate,

rapid, and relatively low-cost monitoring and characterization

methods for the determination of platinum anticancer drugs

has strongly increased in recent years.1

A number of methods are currently available for characterizing

platinum drugs,2–6 including X-ray fluorescence, proton-induced

X-ray emission, inductively coupled plasma (ICP) methods,

flameless atomic absorption, high-performance liquid chromato-

graphy (HPLC), and NMR. However, all these methods suffer

from the crucial drawback of poor sensitivity. Recently, mass

spectrometry (MS) has been applied to detect platinum

complexes owing to this method’s high sensitivity, large linear

dynamic range, speed of response, and ability to accommodate

the reactive, complex nature of the biological samples being

analyzed. However, the conventional MS technique of electron

impact ionization (EI-MS) has proved unsuccessful for

providing useful information about biological samples. This

lack of mass spectral data is due largely to the low volatility of

biological compounds as well as their decomposition at

moderate temperatures (270 1C). For successful analysis of

platinum drugs, MS analytical methods must accommodate

thermally labile, nonvolatile, and highly insoluble nature of

platinum coordination compounds. An intensive molecular

ion has been obtained, that is supposed to significantly

improve the understanding of cisplatin metabolism for the

optimization of chemotherapy.

In contrast, methods comprising evolved gas analysis

coupled with mass spectrometry (EGA-MS) have recently

been developed and implemented in many laboratories.7–10

EGA-MS can be considered as a second-generation version of

pyrolysis mass spectrometry and has been applied to study the

temperature-programmed decomposition of a wide variety of

materials, such as salts, materials relevant to the nuclear

industry, complex salts of d-metals, and polymers. In addition

to a mass spectrum, a total ion monitor (TIM) signal against

temperature, known as a thermogram (pyrogram), can also be

obtained, and such thermograms have been utilized for

constructing decomposition profiles. EGA-IAMS can be also

used as a temperature-programmed reactor for non-isothermal

kinetic studies. Therefore, EGA-MS appears to be promising

for the characterization of thermally labile, nonvolatile, and

highly insoluble materials such as cisplatin.

On the other hand we have shown recently that Li+ ion

attachment mass spectrometry (IAMS) is better than traditional

EI-MS11–15 for the monitoring of the thermally labile

Cu(hfac)(tmvs) found in a copper chemical vapor deposition

(Cu-CVD) reactor compared with the monitoring capabilities

of traditional EI-MS. Unlike the EI-MS, which induces

ionization by means of high-energy electrons, IAMS preserves

aMeisei University, Department of Chemistry, Faculty of Sciences andEngineering, Hodokubo 2-1-1, Hino, Tokyo 191-8506, Japan.E-mail: fujii@chem.meisei-u.ac.jp

bCanon Anelva Technix Corp., 2-5-1 Kurigi, Asao,Kawasaki 215-8550, Japan

3910 | Phys. Chem. Chem. Phys., 2010, 12, 3910–3913 This journal is �c the Owner Societies 2010

PAPER www.rsc.org/pccp | Physical Chemistry Chemical Physics

Publ

ishe

d on

24

Febr

uary

201

0. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 26

/10/

2014

07:

44:3

1.

View Article Online / Journal Homepage / Table of Contents for this issue

the structure of the Cu(hfac)(tmvs) molecules, allowing us to

detect them as adduct ions without any fragmentation. In the

present study, we investigated the potential advantages and

limitations of IAMS as an alternative to conventional EI-MS

for the analysis of platinum drugs.

We have applied a new technology based on Li+ IAMS with

a conventional direct probe and an infrared image furnace

(IIF) to characterize cisplatin. This paper describes the scope

and capabilities of our EGA-IAMS facility as well as

the thermal characterization results of the temperature-

programmed decomposition of cisplatin. We also evaluated

the ability of IAMS to address the aforementioned challenges

associated with cisplatin analysis. Using IAMS, we obtained

the decomposition energy of NH3 from pyrolysis of cisplatin.

We also detected the molecular ion of the Li+ ion adduct of

cisplatin with substantial signal intensity; to our knowledge

this is the first reported instance of cisplatin Li+ ion adducts.

These results suggest that IAMS is suitable for monitoring

platinum drugs with direct, real-time analysis of the metabolite

products, which is expected to significantly improve the

understanding of parent drug and its metabolites.

Experimental

Experiments were performed with an IAMS apparatus (ion

attachment mass spectrometer, IA-Lab) manufactured by

Canon ANELVA Technix Corp (Kawasaki, Japan). This

apparatus features a quadrupole mass spectrometer with a

lithium ion emitter and a direct inlet probe (DIP). The DIP

allows for direct insertion and heating of solid or liquid

samples to prepare them for evaporation under vacuum. The

mass range of this instrumentation extends to m/z 1000. The

mass spectra shown in this paper were obtained under ion

attachment (IA) ionization conditions. Powdered samples of

the cisplatin compound were purchased from Aldrich (99.9%)

and were used as received. The samples were placed in a 2-mm

quartz capillary tube at the end of the DIP and were inserted

into the reaction chamber through a standard airlock. The

background pressure of the reaction chamber prior to sample

introduction was ca. 100 Pa.

A homemade atmospheric pressure infrared image furnace16

(IIF) was coupled with the Li+ IAMS for use in this study.

Detailed construction of this IIF-IAMS system is reported

elsewhere.17 Briefly, a Sinku-Riko IIF (model no. RHL-E45P)

was employed as the heat source for this system. We developed

an orifice interface system to act as an interface between

samples contained in the IIF at atmospheric pressure and

the high-vacuum environment inside the mass spectrometer.

Using this system, we are able to detect any chemical species at

atmospheric pressure, including radical intermediates, by

means of evolved gas analysis.

In our orifice interface system, a powdered sample in a quartz

pan was heated to decomposition in a temperature-programmed

IIF. The decomposition products were swept by a nitrogen

stream at a flow rate of 250–300 ml min�1 and 1 atm pressure,

through an orifice into the mass spectrometer. The temperature

of the orifice region was normally kept above 200 1C.

In kinetic studies, the IIF can be used as a temperature-

programmed flow reactor for drug decomposition. In addition

to the simple Arrhenius equation, the isoconversional method

of Flynn and Wall18 as well as other calculations9,10,19 can be

used to determine the apparent activation energy Ea in such

studies. MS data acquired from multi-ion monitoring serves as

a basis for these calculations.

Results and discussion

Mass spectrum

Fig. 1 shows a typical Li+ adduct mass spectrum of the

as-received reagent [Pt(NH3)2Cl2] for the mass range up to

m/z 350. The m/z values reported here for platinum-containing

ions refer to the 194Pt isotope. This mass spectrum was

acquired with the DIP at 270 1C, a temperature low enough

to allow ionization of the [Pt(NH3)2Cl2] complex without

severe decomposition.

A cluster of peaks corresponding to molecular ions containing

various platinum and chlorine isotopes were observed for Li+

adduct molecular ions, without fragmentation. Eight peaks

were observed at m/z 305, 306, 307, 308, 309, 310, 311 and 313;

these peaks contained platinum isotopes of m/z 194, 195, 196

and 198 and chlorine isotopes of m/z 35 and 37. These peaks

had the expected m/z values and appeared in the ratios

expected according to their natural abundances. In addition

to the observation of the molecular ion, an important feature

of this IAMS spectrum is the presence of peaks at m/z

290 and m/z 324, corresponding to [Pt(NH3)Cl2]Li+ and

[Pt(NH3)3Cl2]Li+, respectively. We believe that these ions

did not result from the fragmentation of [Pt(NH3)2Cl2]Li+,

but from the presence of traces of reagents employed in the

synthesis of cisplatin. A cluster of peaks at m/z 334 also was

observed in the mass spectrum shown in Fig. 1, but we were

unable to identify these peaks. The presence of this unidentified

ionic species further supports our conclusion that the as-received

cisplatin compound contained some residual reagents that

were used in its synthesis.

This ion attachment mass spectrum is completely different

from an electron ionization (EI) mass spectrum. According to

the Riley’s Fourier-transform–MS study,20 a 50-eV EI mass

Fig. 1 An ion attachment mass spectrum of a cisplatin sample heated

to B270 1C. Samples were placed in the IIF and then heated linearly

from room temperature to 600 1C at 20 1C min�1. The computer

simulation spectrum for the molecular ion cluster (calculated peak

intensities, %) are: m/z 305 (74%), m/z 306 (78%), m/z 307 (100%),

m/z 308(49%),m/z 309 (56%),m/z 310 (8%) andm/z 311 (15%). Note;

the m/z 307 for platinum-containing molecular ions corresponds to the

(NH3)2194Pt35Cl37Cl 7Li, (NH3)2

196Pt35Cl35Cl 7Li, and (NH3)2195Pt35Cl37Cl 6Li.

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 3910–3913 | 3911

Publ

ishe

d on

24

Febr

uary

201

0. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 26

/10/

2014

07:

44:3

1.

View Article Online

spectrum of cisplatin gives many fragment ions such as m/z

283 (–NH3), m/z 265 (–Cl or–2NH3), m/z 248 (–Cl and –NH3),

m/z 213 (–2Cl and –NH3), and m/z 196 (Pt). These fragment

ions are attributed to extensive ligand cleavage under harsh EI

conditions. Furthermore, Riley’s group observed clusters

resulting from the addition of one or two chlorines

(Pt(NH3)2Cl3+ and Pt(NH3)2Cl4

+, respectively), whereas we

observed only [Pt(NH3)2Cl3]Li+. No explanation for these

discrepancies can be made at this time.

The HPLC-electrospray mass spectrum21,22 of cisplatin

differed from that of EI and ion attachment ionization, giving

[M � H2O]+, M+, and [M � H2O + MeCN]+ clusters

beginning at m/z 263, 281, and 304, respectively, in addition

to a Na+ adduct ion when acetonitrile was used in the HPLC

solvent mixture.

Temperature-resolved analysis

Fig. 2 shows multi-ion monitoring for NH3Li+ (m/z 24)

and HClLi+ (m/z 43) for a cisplatin sample subjected to

temperature-programmed heating to 620 1C in the IIF. NH3

was detectable from 220 to 370 1C and HCl from 270 to

380 1C. The ion intensities of H2OLi+ (m/z 25) were also

monitored to determine whether water was interfering with the

NH3 analysis. The ratio of H2OLi+ (m/z 25) to NH3Li+

(m/z 24) was less than 0.05 at all temperature. Thus, the

contributions due to background water were negligible.

NH3 and HCl species were evolved simultaneously during

temperature-programmed heating. The ‘‘tail’’ associated with

the m/z 43 response resulted from slow removal of HCl7

from reaction chamber, as evidenced by a persistent, slowly

decreasing m/z 43 background after removal of the cisplatin

sample from the chamber.

Decomposition energy

The IIF-IAMS apparatus described here has been used

previously to study the temperature-programmed decomposition

of several drugs. In this report, we describe some preliminary

experimental results on the temperature-programmed decom-

position of cisplatin.

The decomposition of chemical compounds may consist of

complicated competing reactions. In this study, however, we

used the following simplified assumptions:

da/dt = k(1 � a) (1)

where a is the degree of yield of the decomposed specimen,

k is the rate coefficient, and the reaction is assumed to be

first-order (i.e., a unimolecular decomposition reaction). The

rate coefficient may be approximated by a simple Arrhenius

temperature dependence:

k = e�(Ea/RT) (2)

where A is the frequency factor, Ea is the apparent activation

energy, T is temperature, and R is the gas constant. With the

addition of the heating rate (b = dT/dt), eqn (1) is

transformed into

da/dT = (A/b)e�(Ea/RT)(1 � a) (3)

Mathematically this equation is not easily solved. To

overcome this problem, Knumann et al.19 applied multiple

linear regression analysis to the logarithmic form of eqn (3):

ln(da/dT) = ln(A/b) � Ea/RT + ln(1 � a) (4)

In addition, the degree of yield a has been used in the

calculations. A pyrogram (i.e., a thermogram under multi-ion

detection conditions) obtained from the EGA-IAMS gives the

relative number of decomposition product molecules, thus

indicating the decomposition rate. The degree of conversion

at any temperature T corresponds to the integrated area under

the pyrogram curve between the temperature at the start of the

pyrogram (T0) and T.

The ionic signal (i) acquired from real-time, multiple-ion

detection of chemicals released from thermally decomposed

specimens has been used to obtain a functional form of kinetic

rate expressions.19 From ln[(da/dT)/(1 � a)] vs. 1/T plots, the

Arrhenius parameters (Ea and A) can be determined.

We investigated the intensities of the NH3 signals at m/z

24 over the whole temperature range of 220–370 1C to obtain

the rate expressions for cisplatin decomposition, under the

assumptions that (i) an equilibrium is established between the

thermal decomposition and the detection of the decomposition

products by Li+ IAMS; (ii) a simple correlation exists between

the adduct ion signal, the product NH3 concentrations, and

cisplatin decomposition; and (iii) the overall decomposition

rate of the heated cisplatin sample remains constant from

225 to 249 1C where the degree of conversion of the sample

(fraction of material decomposed) is well-defined. The slope of

the plot of signal intensity versus temperature in this limited

region was constant (Fig. 3). Values of Ea = 38.0 kcal mol�1

and A = 2.2 � 1012 were calculated from this plot.

The pyrolysis of cisplatin is initiated mostly by bond breaking.

Therefore, the above decomposition energy (Ea) may be

associated with the bonding energy that holds the cisplatin

molecule together. Since no experimental value of the bonding

energy is available, preliminary density functional theory

(DFT) calculations23 with an SDD basis set for Pt and a

D95V basis set for the remaining atoms have been performed

for cisplatin. The bonding energy in these DFT calculations

Fig. 2 An EGA curve (selected-ion monitoring) for cisplatin (sample

weight, 1 mg; heating rate, 20 1Cmin�1; atmosphere, nitrogen). Shown

here are relative ion intensities for m/z 24 (NH3Li+) obtained during

temperature-programmed heating of a cisplatin sample to 620 1C in

the IIF.

3912 | Phys. Chem. Chem. Phys., 2010, 12, 3910–3913 This journal is �c the Owner Societies 2010

Publ

ishe

d on

24

Febr

uary

201

0. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 26

/10/

2014

07:

44:3

1.

View Article Online

was obtained as the total energy difference between

[Pt(NH3)2Cl2] and [PtCl2] plus (NH3)2. The DFT calculations

indicate that the bond energy De for Pt–NH3 is 40.4 kcal mol�1.

This value is roughly the same as the decomposition energy for

Pt–NH3 that was obtained in the present study

(Ea = 38.0 kcal mol�1). Note, however, that discrepancy

between our calculated values of Ea and decomposition energy

of Ea might be better explained considering (i) the above

mentioned assumption and (ii) that cisplatin does not behave

as an isolated molecule in the solid.

Concluding remarks

This study has demonstrated the successful application of

IAMS to characterize the drug cisplatin [Pt(NH3)2Cl2]. Our

IAMS system, along with an infrared image furnace, provided

cisplatin molecular ions as Li+ ion adducts [Pt(NH3)2Cl2Li+],

which could be used to monitor volume balances for metabolic

studies of cisplatin compound. To the best of our knowledge,

this is the first reported instance of cisplatin Li+ ion adducts.

This novel IAMS method is promising for the continuous

monitoring of cisplatin in patients.

This study also demonstrates the utility of the combination

of an infrared image furnace (IIF) with IAMS for kinetics

studies. In this study, the decomposition energy of NH3 from

cisplatin pyrolysis was found to be 38.0 kcal mol�1. This

decomposition parameter is important for predicting the

drug’s stability. The combined approach of IIF-IAMS is

promising for future kinetic characterization studies.

Acknowledgements

We would like to express our appreciation to Dr Raja for his

careful analysis of the DFT calculations. This work was

supported in part by a grant from the France-Japan Sasakawa

Foundation (Code 08-PT/6).

References

1 B. Rosenberg, Canc. Treat. Rep., 1979, 63(9–10), 1433–1438.2 W. A. J. De Waal, F. J. M. J. Maessen and J. C. Kraak, J. Pharm.Biomed. Anal., 1990, 8(1), 1–30.

3 T. J. M. Hodes, W. J. M. Underberg, G. Los and J. H. Beijnen,Pharm. Weekbl. Sci., 1992, 14(3), 61–77.

4 S. J. Bannister, L. A. Sternson, A. J. Repta and G. W. James, Clin.Chem., 1977, 23(12), 2258–2262.

5 V. Augey, M. Cociglio, M. Galtier, R. Yearoo, V. Pinsani andF. Bressolle, J. Pharm. Biomed. Anal., 1995, 13(9), 1173–1178.

6 W. A. J. De Waal, F. J. M. J. Maessen and J. C. Kraak,J. Chromatogr., A, 1987, 407, 253–272.

7 M. A. Richard and R. J. Pancirov, J. Therm. Anal., 1987, 32(3),825–834.

8 M. Koel, M. Kaljurand and C. H. Lochmueller, Anal. Chem., 1997,69(22), 4586–4591.

9 S. Dash, P. K. Ajikumar, M. Kamruddin and A. K. Tyagi,Thermochim. Acta, 1999, 334(1–2), 141–148.

10 M. Kamruddin, P. K. Ajikumar, S. Dash, A. K. Tyagi andBaldev Raj, Bull. Mater. Sci., 2003, 26(4), 449–460.

11 T. Fujii, Ion Attachment Mass Spectrometry, ‘‘Encyclopedia ofMass Spectrometry, Ionization Methods’’, ed. Michael Gross,America Society for Mass Spectrometry, Elsevier, 2007, vol. 6,pp. 327–334.

12 M. Sablier and T. Fujii, Annu. Rep. Prog. Chem. Sect. C: Phys.Chem., 2005, 101, 53–99.

13 T. Fujii, Chem. Phys. Lett., 1992, 191, 162.14 T. Fujii, Phys. Rev. E: Stat. Phys. Plasmas Fluids, 1998, 58(5–B),

6495–6502.15 T. Fujii, J. Appl. Phys., 1997, 82, 2056.16 M. J. Davis and P. D. Ihinger, Am. Mineral., 1999, 84(1–2),

48–54.17 Y. Kitahara, S. Takahashi, N. Kuramoto, M. Sala, T.

Tsugoshi, M. Sablier and T. Fujii, Anal. Chem., 2009, 81(8),3155–3158.

18 J. H. Flynn and L. A. J. Wall, J. Polym. Sci., Part B: Polym. Lett.,1966, 4, 323.

19 R. Knumann and H. Bockhorn, Combust. Sci. Technol., 1994,101(1), 285–299.

20 R. R. Weller, J. R. Eyler and C. M. Riley, J. Pharm. Biomed. Anal.,1985, 3(1), 87–94.

21 H. C. Ehrsson, I. B. Wallin and S. Anita, Anal. Chem., 1995,67(19), 3608–3611.

22 M. Cui and Z. Mester, Rapid Commun. Mass Spectrom., 2003, 17,1517–1527.

23 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,M. A. Robb, J. R. Cheeseman, J. A. Montgomery, T. Vreven,K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi,V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,G. A. Petersson, H. Nakatsuji, M. E. M. Hada, K. Toyota,R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma,G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski,S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas,D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman,J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. ioslowski,B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi,R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,B. Johnson, W. Chen, M. W. Wong, C. Gonzalez andJ. A. Pople, g03RevB version ed., Gaussian, Inc., WallingfordCT, 2004.

Fig. 3 An Arrhenius plot for NH3 produced from cisplatin decom-

position in the temperature range from 225 to 249 1C.

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 3910–3913 | 3913

Publ

ishe

d on

24

Febr

uary

201

0. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 26

/10/

2014

07:

44:3

1.

View Article Online