Entertainment & Humor

A robust method to quantify low molecular weight contaminants in heparin: detection of tris(2-n-butoxyethyl) phosphate

Recently, oversulfated chondroitin sulfate (OSCS) was identified in contaminated heparin preparations, which were linked to several adverse clinical events and deaths. Orthogonal analytical techniques, namely nuclear magnetic resonance (NMR) and
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A robust method to quantify low molecular weight contaminants in heparin:detection of tris(2- n -butoxyethyl) phosphate † Guilherme L. Sassaki, * a Daniel S. Riter, a Arquimedes P. Santana Filho, a Marco Guerrini, b Marcelo A. Lima, c  Cesare Cosentino, b Lauro M. Souza, a Thales R. Cipriani, a Timothy R. Rudd, bd  Helena B. Nader, c  Edwin A. Yates, d  Philip A. J. Gorin, a Giangiacomo Torri b and Marcello Iacomini a Received 16th December 2010, Accepted 23rd March 2011 DOI: 10.1039/c0an01010c Recently, oversulfated chondroitin sulfate (OSCS) was identified in contaminated heparinpreparations, which were linked to several adverse clinical events and deaths. Orthogonal analyticaltechniques, namely nuclear magnetic resonance (NMR) and capillary electrophoresis (CE), have sincebeen applied byseveral authors for the evaluation of heparin purity and safety. NMR identification andquantification of residual solvents and non-volatile low molecular contaminants with USP acceptancelevels of toxicity was achieved 40-fold faster than the traditional GC-headspace technique, which takes  120 min against  3 min to obtain a  1 H NMR spectrum with a signal/noise ratio of at least 1000/1. Theprocedure allowed detection of Class 1 residual solvents at 2 ppm andquantification was possible above10 ppm. 2D NMR techniques (edited-HSQC  1 H/ 13 C) permitted visualization of otherwise maskedEDTA signals at 3.68/59.7 ppm and 3.34/53.5 ppm, which may be overlapping mononuclear heparinsignals, or those of ethanol and methanol. Detailed NMR and ESI-MS/MS studies revealed a hithertounknown contaminant, tris(2- n -butoxyethyl) phosphate (TBEP), which has potential health risks. 1. Introduction Heparin, a glycosaminoglycan (GAG), is the most widelyemployed anticoagulant and antithrombotic drug in medicine.During the past 3 years, an oversulfated chondroitin sulfate(OSCS) contaminant has been isolated from heparin prepara-tions, 1 to which were attributed several clinical manifestations,including hypotension and direct activation of the kinin-kalli-krein pathway in human plasma leading to the generation of bradykinin, a potent vasoactive mediator. In addition, OSCSinduces generation of C3a and C5a, potent anaphylatoxinsderived from complement proteins. 2,3 The activation of these twopathways was unexpectedly linked to, and dependent on, fluid-phase activation of factor XII, which led to the death of at least238 patients. 3,4 These oversulfated GAGs also had anticoagulantactivity, as well as the capability of activating the contact system.Together, these events led to the world wide heparin crisis in2008, 4–6 which has stimulated the application of orthogonalmethods, including NMR, capillary electrophoresis, HPLC andthe use of principal component analysis, to identify potentialcontaminants in heparin preparations. 1,7–13 Since the OSCScontamination event, heparin preparations have to be analyzedby  1 H NMR spectroscopy, following USP and European phar-macopeia manuscripts. 14,15 While OSCS is a contaminant, itshould be noted that dematan sulfate (DS) is an acceptedimpurity. Low molecular weight components such as ethyl-enediamine tetraacetic acid (EDTA) and hazardous residualsolvents have also been detected using specific regions of theirNMR spectra. Residual contaminants in pharmaceutical prod-ucts are often volatile organic compounds, typically solventsused during their manufacture which have not been removedcompletely. 14 The classification for residual solvents is based onthe toxicity of each possible chemical component, and fall intothree classes: Class 1 -  Solvents to be avoided   (carcinogenic,strongly suspected human carcinogens and environmentalhazards); Class 2 -  Solvents to be limited   (nongenotoxic animalcarcinogens or possible causative agents of other irreversibletoxicity, such as neurotoxicity or teratogenicity and suspected of other significant, but reversible, toxicities) and Class 3 -  Solventswith low toxic potential   (no health-based exposure limit is deemednecessary). These compounds are typically detected and quan-tified using chromatographic techniques, such as gas a Departamento de Bioqu  ımica e Biologia Molecular, Universidade Federal do Paran  a, Curitiba, PR CEP: 81531-980, Brazil  b Istituto di Richerche Chimiche e Biochimiche ‘‘G. Ronzoni’’, via GiuseppeColombo 81, 20133 Milan, Italy c Departamento de Bioqu  ımica e Biologia Molecular, Universidade Federal de Sa˜o Paulo, SP 04044020, Brazil  d  School of Biological Sciences, University of Liverpool, Liverpool, L697ZB, UK  † Electronic supplementary information (ESI) available: S1: Table 1Values of integrated areas of the contaminant/H-1 heparin ratio, S2:Table 2 Values of sum from signal height contaminant/H-1 heparinratio, S3: PCA analysis from  1 H-NMR spectra of heparin samples atdifferent contaminant concentrations. See DOI: 10.1039/c0an01010c 2330 |  Analyst  , 2011,  136 , 2330–2338 This journal is  ª  The Royal Society of Chemistry 2011 Dynamic Article Links C < Analyst  Cite this:  Analyst  , 2011,  136 , 2330www.rsc.org/analyst  PAPER    P  u   b   l   i  s   h  e   d  o  n   1   4   A  p  r   i   l   2   0   1   1 .   D  o  w  n   l  o  a   d  e   d   b  y   U  n   i  v  e  r  s   i   t  y  o   f   W  a  s   h   i  n  g   t  o  n  o  n   1   2   /   0   6   /   2   0   1   3   1   9  :   5   7  :   3   2 . View Article Online / Journal Homepage / Table of Contents for this issue  chromatography, between specific limits according to theirclassification. 16 The most common residual solvents found inheparin preparations are ethanol, methanol, acetic acid, acetone,phenol, and benzyl alcohol and their acceptable limits rangefrom 2 to 5000 ppm. 16 Here, we report a robust NMR identifi-cation and quantification method applied to the detection of common heparin residual solvents and some low molecularcontaminants found in heparin preparations. The approachcombines rapid identification and quantification of residualsolvents at acceptable limits, as well as other contaminants, bymeans of a single  1 H NMR spectrum, ensuring purity andsubsequent human safety, making it an alternative orthogonalmethod to traditional gas chromatography assays. Furthermore,the structure of a hitherto identified contaminant present in someBrazilian heparin preparations that uses Chinese raw materialhas been identified by detailed NMR and ESI-MS/MS studies astris(2- n -butoxyethyl) phosphate (TBEP). 2. Experimental 2.1 Chemicals Heparin, sodium salt (porcine and bovine intestinal mucosa),chemicals (EDTA and phenol) and residual solvents (methanol,ethanol, benzyl alcohol and acetic acid/sodium acetate) werepurchased from Sigma-Aldrich, MO, USA, MERCK (Darm-stadt, Germany) and from European Directorate for the Qualityof Medicines and HealthCare (EDQM). Twenty two heparinpreparations were obtained from various pharmaceuticalcompanies and seven heparin vials were purchased from theBrazilian market. Samples (20 mg) of pure bovine heparin andmixtures of residual solvents at concentrations from 0.015–2%,w/w, were prepared by dissolving the sample in 0.5 mL D 2 O(99.97%, Sigma-Aldrich, MO, USA). 2.2 Gas chromatography identification and quantification of residual solvents Gas chromatography (GC) was performed, following the USP467 monograph 16 in two Brazilian laboratories certified byANVISA (Ag ^ encia Nacional de Vigil ^ ancia Sanit  aria). The GCsystem was equipped with a headspace injector and a flame-ionization detector, a fused-silica 0.32 mm    30 m columncoated with a 1.8 mm layer of phase G43, or a 0.53 mm  30 mwide-bore column coated with a 3.0 mm layer of phase G43. Thecarrier gas was nitrogen with a linear velocity of about 35 cm s  1 and a split ratio of 1/5. The column temperature was maintainedat 40   C for 20 min, then raised at 10   C min  1 to 240   C, thenheld at 240   C for 20 min. The injection port and detectortemperature were held at 140   C and 250   C respectively, and theconcentrations of ethanol and methanol were determined and areexpressed in ppm. 2.3 NMR spectroscopy NMR was performed using a Bruker 400 MHz or 600 MHzAVANCE  III   NMR spectrometer (Bruker GmbH, Silberstrei-fen, Germany) with a 5 mm inverse Z gradient probe or TCIcryoprobe.  1 H-NMR and  13 C-NMR spectra of the heparinpreparations were obtained after dissolving the samples in D 2 O,without freeze-drying to avoid the residual solvent evaporation.The chemical shifts of signals were measured relative to themethyl group of the N-acetyl of a Glc-NAc standard at 303 K(24.3 and 2.050 ppm for  13 C and  1 H signals, respectively). 1D  1 H-NMR was performed, as described in the USP and EuropeanPharmacopeias relating to heparin monographs, 14,15 providinga sensitive method in terms of limits of detection (LOD), thespectra were acquired using 16 scans to give a signal/noise (S/N)ratio of at least 1000/1 (90  pulse, relaxation delay  ¼  4.0 s,number of time domain points ¼ 65536, and acquisition time ¼ 7.7 s). In general, experiments were performed without tuberotation with the HOD signal at a medium width varying from 2– 3.5 Hz. 2D NMR experiments were carried out using edited-HSQC (hsqcedetgp), COSY (cosygpprqf), TOCSY (mlevphpr.2),and HMBC (hmbcgplpndqf) pulse sequences, recorded forquadrature detection in the indirect dimension and acquiredusing 8 scans per series of 1024    512 data points, with zerofilling in F1 (4096) prior to Fourier transformation. 2.4 Isolation and Identification of the low molecularcontaminant The contaminated commercial heparin lot (vial 5 mL) wastransferred to a borosilicate flask with teflon screw caps andpartitioned (  3) with CHCl 3  (5 mL). The organic layer wasevaporated atroom temperature in a fumehood. The residue (400 m g) was dissolved in CDCl 3  –MeOD (500  m L 5 : 0.1; v/v), to shiftthe residual water peak to an appropriate position and analyzedby NMR and ESI-MS/MS. 2.5 ESI-MS/MS analysis A sample of the TBEP contaminant (10 m g mL  1 ) was solubilizedin MeOH–H 2 O (7 : 3, v/v) containing 5 mM LiCl, and submittedto positive and negative API-ESI-MS ionization analysis,recorded using triple quadrupole Quattro LC (Waters, Milford,USA) equipment, with nitrogen as nebulizer and desolvation gas.Offline analysis was performed by direct sample injection of methanol solutions into the ESI-MS source with a syringe-infu-sion pump at a flow rate of 10  m L min  1 . Second stage tandem-MS profiles were obtained by collision induced dissociation(CID), using helium as collision gas. Positive ions were obtainedwith capillary (3.3 kV), cone (81 V), and collision energy (77 eV). 2.6 Multivariate analysis 1 H-NMR spectra were saved as ASCII files  via  Mnova NMR6.1.1 (Mestrelab Research Company) for further PCA andcluster analysis, which were performed using software R (R: ALanguage and Environment for Statistical Computing; RFoundation for Statistical Computing, Vienna, Austria. http://cran.r-project.org/), with prior mean centering. 3. Results 3.1 Assembling the curve of low molecular weight contaminants Heparin, residual solvents, and phenol samples were dissolved inD 2 O. The contaminant concentrations ranged from 155 to 20000ppm, based on values of ‘‘tolerable daily intake’’ (TDI) to This journal is  ª  The Royal Society of Chemistry 2011  Analyst  , 2011,  136 , 2330–2338 | 2331    P  u   b   l   i  s   h  e   d  o  n   1   4   A  p  r   i   l   2   0   1   1 .   D  o  w  n   l  o  a   d  e   d   b  y   U  n   i  v  e  r  s   i   t  y  o   f   W  a  s   h   i  n  g   t  o  n  o  n   1   2   /   0   6   /   2   0   1   3   1   9  :   5   7  :   3   2 . View Article Online  describe exposure limits of toxic chemicals, ‘‘acceptable dailyintake’’ (ADI) and ‘‘permitted daily exposure’’ (PDE). Valuesabove 5000 ppm were used to test NMR quantification limits. Toguarantee the robustness and linearity of the method, integrationof the H-1 region from 5.70 to 4.90 ppm of the spiked heparinsamples was carried out, since variations in GlcUA, IdoUA(2 S  ),GlcNAc, GlcNS, GlcNAc(6 S  ), and GlcNS(6 S  ) content arecommonly found in commercial heparin preparations. This alsoavoids the influence of the water peak on the surrounding signals.We also compared the signal intensity of the contaminants andH-1 heparin signals. A direct ratio of contaminant/H-1 heparinsignals was calculated and correlated to each contaminantconcentration (Fig. 1) (ESI S1†). The resulting data showed thatthe sensitivity was not identical for each contaminant, since smalldifferences were observed for the C H  3  singlet and triplet frommethanol and ethanol respectively, the better response beingobserved for ethanol, since it is not superimposed, unlike themethanol singlet, which appears together with G-2 residues of heparin (Fig. 1). Linear regression of the data from thecontaminants showed satisfactory values for the correlationcoefficient ( R 2 ) varying from 0.974 to 0.997, although the S/NNMR response, concentration, and transients did not followa linear function, besides the exponential function gave rise to  R 2 varying from 0.997 to 1.000 (Fig. 1) (ESI S1†). Similar resultswere obtained to the experiment carried out with the sum of values from signal intensity of the contaminants and H-1heparin signals (Fig. 1, ESI S2†). For both methods, standarddeviations and correlation values were validated with  p  < 0.001(ESI S1 and S2†). The  1 H NMR spectra also showed thatsolvents and low molecular contaminants could be detected atlow concentrations, as demonstrated in an expanded spectrumfrom a sample containing 155 ppm of low molecular weightcontaminants in 20 mg of bovine intestinal mucosa heparin(Fig. 1). In order to confirm our results, PCA analysis wascarried out. 17,18 The first two components covered 98.22% of thevariance within the dataset, PC1 (95.76%) and PC2 (3.71%),being used to differentiate the samples by the concentration of residual solvents and phenol (non-volatile) ranging from 155 to20000 ppm [ESI S3† (Fig. 1A)]. The presence of the solventcontaminants produced unique spectra, which were readilydifferentiated by PCA of their respective NMR spectra andeasily observed by hierarchical cluster analysis [ESI S3†(Fig. 1B)]. PC1 and PC2 score plots illustrate that the maindifferences are related to the contaminant peaks and not to theheparin signals [ESI S3† (Fig. 1C and D)]. These results areexpected, since the only modification carried out was thechange in contaminant concentration. The multivariate analysisshowed that NMR quantification was reproducible and robust,and exhibiting less interference, when compared with the headspace GC system (Table 1). Fig. 1  (A)  1 H-NMR spectra and their specific regions used for integration, H-1 of bovine heparin and those containing contaminants at differentconcentrations. Abbreviations of each contaminant follow IUPAC recommendations: benzyl alcohol (BzOH), phenol (PhOH), methanol (MeOH),acetate (AcO  ), and ethanol (EtOH) varying from 155 to 20000 ppm. (B) Calibration curves obtained of integrated regions from contaminant/H-1heparin at different concentrations (ppm) for each contaminant; linear regression showed  R 2 values varying from 0.974 to 0.997. (C) Calibration curvesobtained of sum from signal height of regions from contaminant/H-1 heparin at different concentrations (ppm) for each contaminant; linear regressionshowed  R 2 values varying from 0.997 to 0.999. 2332 |  Analyst  , 2011,  136 , 2330–2338 This journal is  ª  The Royal Society of Chemistry 2011    P  u   b   l   i  s   h  e   d  o  n   1   4   A  p  r   i   l   2   0   1   1 .   D  o  w  n   l  o  a   d  e   d   b  y   U  n   i  v  e  r  s   i   t  y  o   f   W  a  s   h   i  n  g   t  o  n  o  n   1   2   /   0   6   /   2   0   1   3   1   9  :   5   7  :   3   2 . View Article Online  3.2 Direct measurement of residual solvents in commercialheparin samples Heparin samples (20 mg) were directly dissolved in D 2 O andanalysed as described above. Integration of the H-1 region fromuncontaminated heparins and those tainted with ethanol andmethanol were performed and the results were compared withthose obtained by gas chromatograph USP (467). 16 Direct  1 H-NMR analysis of all heparin samples showed ethanol andmethanol at different concentrations. We selected threecommercial heparins commonly used in hemodialysis andcardiovascular surgery and they showed benzyl alcohol, whichis utilized as an antiseptic in heparin preparations (  60000ppm). These values are twelve-fold greater for Class 3 residualsolvents. Quantification of other contaminants showed ethanoland methanol at 420 and 3630 ppm, respectively (Fig. 2;porcine (1) and bovine heparin (2) preparations), the latterbeing greater than permitted limits for Class 2 solvents. Wealso observed the presence of DS in most of the heparinpreparations (Fig. 2). However, the main differences wereobserved for commercial heparin (Fig. 2 (spectrum 3)), sincethis preparation appears to contain a bivalent ion ( e.g. calcium). The spectrum obtained after EDTA addition showedthe typical profile of porcine intestinal mucosal heparin, con-firming this impurity (Fig. 2, spectrum 3a). The C H  3  signals inthe N-acetyl region also showed the presence of other uniden-tified GAGs or contaminants (Fig. 2, spectrum 3b). Curiously,weak aliphatic signals appeared at 0.83, 1.27 and 1.46 ppm inthe spectra of commercial heparin. To identify these signals inolder samples from the same manufacturer a batch from 2007was analysed. The same signals were present, as well as acetateanions at a concentration of 238 ppm. Signals assignable to Table 1  Average area of the contaminants by GC headspace system b Contaminants ppm a 20000 10000 5000 2000 1000 RSD  R 2 Ethanol - C3 301.2 159.0 82.7 45.1 22.0 7.2 0.999Methanol - C2 209.7 106.2 55.3 28.3 14.2 6.8 0.999 a Samples were diluted 100 fold to reach linearity.  R 2 of linearized datafrom the average area and average % relative standard deviations(RSD) were calculated from the four replicate run of each solvent. GCanalysis was performed on a DB 624 (G43 equivalent phase) columnunder USP 467 monograph conditions.  b C2, Class 2, solvents to belimited; C3, Class 3, solvents with low toxic potential. Fig. 2  (1,2,3)  1 H-NMR spectra from amplified portions of different regions from commercial heparins; (3a) commercial heparin (3) with EDTAaddition; (3b) commercial heparin (3), from 2007; (3c) organic fraction after CHCl 3  –H 2 O partition from commercial heparin 3. A-1 to A-5 (correspondsto H1-6 of 6- and  N  -disulfated  a -glucosamine units); I-1 to I-5 (corresponds to H1-5 of 2-sulfated  a -iduronic acid units); B-1 (to H-1 of 6-sulfated  N  -acetyl a -glucosamine units);C-1 (corresponds to H1 of 6-DeS-NAc- a -glucosamineunits); G-2 (to H-2 of  b -glucuronic acidunits); I-1a; I-1b; I-1c (to H-1of   a -iduronic units neighboring 6/ N  -disulfated, 6-sulfated and  N  -acetylated or  N  -sulfated  a -glucosamine units, respectively). Oversulfated condroitinsulfate (OSCS); oversulfated dermatan sulfate (OSDS); dermatan sulfate (DS); contaminants (Conts); heparin (Hep); benzyl alcohol (BzOH); acetate(AcO  ); ethylenediamine tetraacetic acid (EDTA). This journal is  ª  The Royal Society of Chemistry 2011  Analyst  , 2011,  136 , 2330–2338 | 2333    P  u   b   l   i  s   h  e   d  o  n   1   4   A  p  r   i   l   2   0   1   1 .   D  o  w  n   l  o  a   d  e   d   b  y   U  n   i  v  e  r  s   i   t  y  o   f   W  a  s   h   i  n  g   t  o  n  o  n   1   2   /   0   6   /   2   0   1   3   1   9  :   5   7  :   3   2 . View Article Online  OSCS (2.16 ppm), DS (2.08 ppm) and OSDS (2.12 ppm) werealso observed. Other contaminants were also present whichmight have given rise to potential adverse effects, thereforea partition of heparin (3) between CHCl 3  and H 2 O was per-formed. The  1 H-NMR spectrum of the organic layer showed thepresence of benzyl alcohol, with its typical aliphatic signals andthree substituted C H  2 s, which were masked by the heparinsignals before partition (Fig. 2, spectrum 3c). 3.3 Preventive detection of Class 1 and Class 2 solvents used onoversulfated side-stream GAGs and intentionally addeddangerous substances The resolution and greater sensitivity, obtained with  1 H-NMRspectroscopy, using low amounts of heparin preparations,encouraged us to carry out the USP recommendations on theother Class 1 and 2 residual solvents. These would be involvedwith oversulfated GAGs, as observed in worldwide heparinpreparations, and also traces of EDTA, which have been asso-ciated with an increase in bleeding effects. 17 In order to obtaina limit of detection (LOD) of 2 ppm (established limit forbenzene) for class 1 residual solvents, 20 mg of commercialheparin were combined with class 1 solvents and the non-volatilecontaminants, phenol and EDTA. In order to obtain this LODfor benzene, the  1 H-NMR spectrum was recorded with pre-saturation and 128 scans to detect benzene as well as other class 1contaminants. However, their quantification was not possiblebecause the S/N ratio was insufficient to give a precise result.Consequently, using 10 ppm of contaminants, quantification wasperformed over a short analysis period.Intentional contaminants, such as oversulfated side-streamGAGs, present in heparin preparations, could be observed by 1 H-NMR, as demonstrated by several researchers. 1,7–11,20 However, at low concentrations, these species are difficult todetect. It is known that NMR signals from low molecular weightspecies are more sensitive and their signals are better resolved.Since oversulfated side-stream GAGs preparations use some lowmolecular weight solvents, such as pyridine and DMF, an anal-ysis was conducted of mixtures containing heparin, DMF (USPlimits, 200 ppm) and pyridine (USP limits, 200 ppm), which areeasily observable with NMR, preventing any intentionalcontamination by synthetic OSGAGs. These substances wereresidues observed after oversulfation of GAGs and have beenused as a sulfate donor (sulfur trioxide-pyridine complex) withDMF as solvent (Table 2 and Fig. 3). 3.4 Isolation and identification of a new low molecular weightcontaminant The isolated product obtained from the organic partition phase(80  m g mL  1 ) was exhaustively evaporated under reducedpressure. The resulting residue was submitted to detailed NMRstudies using homonuclear (COSY and TOCSY), heteronuclear(edited-HSQC), and heteronuclear multiple bond correlationHMBC { 1 H; 13 C} and { 1 H; 31 P} techniques. The number of hydrogen atoms directly linked to carbon atoms was deter-mined by edited-HSQC, which showed the presence of sevencross peaks and gave six negative  C  H 2  and one positive signalof   C  H 3  (Fig. 4A). COSY and TOCSY showed that aliphatichydrogens have connectivity along the structure of thecontaminant and one of them is a substituted  C  H 2  at 71.3/3.38ppm, namely an H3 (Fig. 4B). The H1 at 4.09 ppm and H2 at3.54 ppm showed connectivity, and no other correlation wasobserved  via  homonuclear experiments (Fig. 4C). Long rangecorrelations { 1 H; 13 C} were observed between the H3 at  d  3.38ppm with carbons at 66.9, 31.7, 19.2 and 14.2 ppm, respec-tively, suggesting that H3 and H1 belong to the same molecule.This was confirmed by correlation of the proton H4 at 1.46ppm with carbon at 69.4 ppm (H2). The 2D NMR data sug-gested the presence of an oxygen atom between H2 and H3nuclei because their carbon chemical shifts are typical of anether linkage, giving rise to  13 C resonances at 69.4 and 71.3ppm, respectively. However, the H1 ( 13 C/ 1 H) resonance at 66.9/4.09 ppm was not readily assignable, since it does not corre-spond to a CH 2  group substituted by amine, hydroxyl or etherbond. Generally, phosphate esters give small  a -shifts, whichenabled detection of phosphate ester linkages in the contami-nant structure. Its  31 P-NMR spectrum showed a single peak at  0.40 ppm which suggests a  31 P nucleus typical of phosphate.In order to confirm this  O -substituent at H1, an HMBC{ 1 H; 31 P} analysis was performed, confirming the phosphateester linkage, with a strong cross peak at   0.40/4.094 ppm anda weak one at   0.40/3.54 ppm (Fig. 4D). The overall NMRresults thus suggest an organophosphate ester and definitiveidentification of contaminant as an organophosphate witha specific structure was provided by ESI-MS analysis. Thesample was solubilized in methanol–H 2 O (7 : 3, v/v) containing5 mM LiCl, and examined in the positive ESI mode usinga capillary energy of 3.3 kV and cone voltage of 81 V. Thisprovided a pseudo-molecular ion [M + Li] + at  m / z  405, whichcorresponds to tris(2- n -butoxyethyl) phosphate (TBEP). TheMS/MS spectrum was obtained at 77 eV, giving second stage Table 2  1 H/ 13 C chemical shifts observed in intentionally contaminated heparin preparations a Heparin/MixContaminants H-1 (aryl) H-2 (aryl) H-3 (aryl) H-4 (aryl) H-1 (CH 3 ) H-2 (CH 3 ) (CH 2 )-NH- COO-(CH 2 )Benzene 7.41/126.9Pyridine 8.49/150.2 7.42/127.1 7.95/141.0Phenol 6.96/123.0 7.32/132.4 6.92/117.8DMF 2.99/39.2 2.83/33.6EDTA 3.68/59.7 3.34/53.5 a Chemical shifts of protons and carbons are expressed in ppm at 303 K, referenced to external TSP. 2334 |  Analyst  , 2011,  136 , 2330–2338 This journal is  ª  The Royal Society of Chemistry 2011    P  u   b   l   i  s   h  e   d  o  n   1   4   A  p  r   i   l   2   0   1   1 .   D  o  w  n   l  o  a   d  e   d   b  y   U  n   i  v  e  r  s   i   t  y  o   f   W  a  s   h   i  n  g   t  o  n  o  n   1   2   /   0   6   /   2   0   1   3   1   9  :   5   7  :   3   2 . View Article Online
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks