Qualitative Analysis of Tackifier Resins in Pressure Sensitive

Aug 26, 2011 - Beiersdorf AG, Unnastrasse 48, D-20245 Hamburg, Germany. ‡. Institute of Organic Chemistry, Department of Chemistry, Faculty of Scien...
0 downloads 0 Views 1MB Size
Download PDF
ARTICLE pubs.acs.org/ac

Qualitative Analysis of Tackifier Resins in Pressure Sensitive Adhesives Using Direct Analysis in Real Time Time-of-Flight Mass Spectrometry Aylin Mess,† Jens-Peter Vietzke,*,† Claudius Rapp,† and Wittko Francke‡ † ‡

Beiersdorf AG, Unnastrasse 48, D-20245 Hamburg, Germany Institute of Organic Chemistry, Department of Chemistry, Faculty of Sciences, University of Hamburg, Martin-Luther-King-Platz 6, D-20146 Hamburg, Germany ABSTRACT: Tackifier resins play an important role as additives in pressure sensitive adhesives (PSAs) to modulate their desired properties. With dependence on their origin and processing, tackifier resins can be multicomponent mixtures. Once they have been incorporated in a polymer matrix, conventional chemical analysis of tackifiers usually tends to be challenging because a suitable sample pretreatment and/or separation is necessary and all characteristic components have to be detected for an unequivocal identification of the resin additive. Nevertheless, a reliable analysis of tackifiers is essential for product quality and safety reasons. A promising approach for the examination of tackifier resins in PSAs is the novel direct analysis in real time mass spectrometry (DART-MS) technique, which enables screening analysis without time-consuming sample preparation. In the present work, four key classes of tackifier resins were studied (rosin, terpene phenolic, polyterpene, and hydrocarbon resins). Their corresponding complex mass spectra were interpreted and used as reference spectra for subsequent analyses. These data were used to analyze tackifier additives in synthetic rubber and acrylic adhesive matrixes. To prove the efficiency of the developed method, complete PSA products containing two or three different tackifiers were analyzed. The tackifier resins were successfully identified, while measurement time and interpretation took less than 10 mins per sample. Determination of resin additives in PSAs can be performed down to 0.1% (w/w, limit of detection) using the three most abundant signals for each tackifier. In summary, DART-MS is a rapid and efficient screening method for the analysis of various tackifiers in PSAs.

P

ressure sensitive adhesives (PSAs) are produced all over the world for various purposes. They are present in adhesive tapes used in the electronics industry and the automotive sector and are used for packaging materials and for healthcare products (sticking plaster). Because of the large number of requirements for different applications, PSAs are very complex formulations containing a large variety of additive components, which enable the product to perform specialized functions. These additives include tackifier resins, plasticizers, cross-linkers, antioxidants, and many others.1 The main polymers used in PSA manufacturing are acrylic adhesive, natural rubber, and synthetic rubber. Especially, rubber elastomers require the addition of tackifier resins to generate adhesive properties. Tackifier resins represent amorphous compounds with relatively low molecular weight and can be divided into four main groups: hydrocarbon resins, polyterpene resins, terpene phenolic resins, and rosin resins. Hydrocarbon resins derive from various petrochemical processes and are mixtures of aromatic and/or aliphatic substances. Polyterpene resins are produced by polymerization of one or several terpenes of plant origin (citric fruits or natural turpentine from trees). By copolymerization of terpene monomers with phenol, so-called terpene phenolic resins are produced, which possess a more hydrophilic character. Similar to terpene-based resins, rosin resins are also obtained from trees r 2011 American Chemical Society

(pines). The main components are abietic acid and other resin acids. As 90% of the latter are isomers of abietic acid, they cannot be distinguished by mass spectrometry.2 Tackifier resins can be modified by hydrogenation, esterification (e.g., rosin-related glycerol ester), and other reactions to influence their properties. Typical amounts of tackifiers in PSAs vary in the range of 10 50% (w/w). It is important to know that abietic acid in rosin resins may cause contact dermatitis and sensitization when applied to human skin.3 Especially, the use of rosin resins in sticking plasters can be unpleasant for people with sensitive skin and who tend to develop allergic reactions. Since the actual composition of PSA formulations is usually unknown, the identification of additives and other components using analytical techniques is becoming more and more important. Moreover, the analysis of tackifiers and additives in PSAs is essential in quality and product safety control. PSAs can be analyzed by Fourier transform-infrared (FT-IR) spectroscopy.4 This is a nondestructive, fast, and relatively simple analytical method which provides simultaneous detection of Received: May 5, 2011 Accepted: August 24, 2011 Published: August 26, 2011 7323

dx.doi.org/10.1021/ac2011608 | Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry several components of a mixture. However, identification of unknown substances is hampered by interference with signals of the polymer matrix, a fact that complicates the analysis of additives like tackifiers, e.g., the acrylic ester band interferes with the ester band of a rosin ester. Extraction of PSA samples to reduce matrix effects might be a good technique for rubber adhesives, but in the case of acrylic adhesives, low molecular weight fragments of the polymer are also soluble in commonly used solvents, so that separation of polymer and tackifier resin is impossible. FT-IR spectroscopy is also insensitive to minor components in mixture analyses, which causes problems as minor amounts of additives are often present. Moreover, the similarity and overlap of additive absorption bands limit the potential of FT-IR spectroscopy as some related additives cannot be distinguished from each other. Another approach in the analysis of PSAs and tackifier resins is pyrolysis-gas chromatography/mass spectrometry (PyGC/MS).4,5 The samples are decomposed during a thermal degradation process, and resulting fragments can be analyzed by GC/MS. This method requires only a short sample preparation time and provides results with high information content. Nevertheless, a chromatographic separation step is needed, which implies timeconsuming analyses. Furthermore, the identification of additives is complicated as fragmentation does not always occur in a selective way. PSAs and tackifier resins have also been analyzed by highperformance liquid chromatography (HPLC),2 capillary electrophoresis,6 ultraperformance liquid chromatography in combination with time-of-flight mass spectrometry (UPLC-TOF-MS),7 and other methods.4,8,9 Recently, the analysis of tackifiers in PSAs using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) has been reported. However, this method was developed only for rosin resins and its derivatives, not for other classes of tackifiers, and sample preparation including extraction, sonication, and centrifugation was necessary.10 In summary, the analysis of tackifier resins in PSAs represents a significant analytical challenge. Because of their composition and huge differences in raw materials and the resulting commercial products, analysis by use of a single specific key component is often impossible. Moreover, the complex and inaccessible polymer matrix causes difficulties in tackifier resin detection and needs sophisticated sample preparation with extraction, purification, and chromatographic separation. Therefore, a screening technique is required which is characterized by good sensitivity, simple sample pretreatment, and the ability to obtain in-depth analytical information. The direct analysis in real time (DART) ion source provides an open atmospheric pressure interface which enables direct introduction of solid, liquid, and gaseous samples without sample preparation.11 Analytes are desorbed from surfaces and undergo a soft ionization process based on the formation of ionized water clusters followed by proton transfer reactions. Attached to a TOF-detector, this experimental apparatus is particularly suitable for wide-range screening analyses with good sensitivity and high mass accuracy, while measurement time only takes a few minutes. DART-TOF-MS has been widely used for the analysis of drugs,12 14 pharmaceuticals,15 21 writing inks,22 explosives, and chemical warfare agents23 25 and for the detection of fungicides in wheat.26 Only a few papers have reported the application of DART for the analysis of polymers,27 30 and it has not yet been applied to PSAs and tackifiers. The aim of this study was to explore the potential of this method for the analysis of tackifier resins in PSA formulations.

ARTICLE

Table 1. Types and Characteristics of the Resins Studied in this Paper sample

type

characteristics

resin 1

rosin resin

glycerol ester, hydrogenated

resin 2 resin 3

rosin resin rosin resin

nonesterified, nonhydrogenated nonesterified, hydrogenated

resin 4

rosin resin

nonesterified, disproportionated

resin 5

rosin resin

pentaerythritol ester, hydrogenated

resin 6

terpene phenolic resin

polymerized α-pinene and phenol

resin 7

polyterpene resin

polymerized β-pinene

resin 8

hydrocarbon resin

polymerized hydrogenated

resin 9 resin 10

hydrocarbon resin hydrocarbon resin

aliphatic C5 type poly(α-methylstyrene)

dicyclopentadiene (DCPD)

’ EXPERIMENTAL SECTION Materials and Reagents. Toluene (HPLC grade), methanol (LC MS hypergrade), dichloromethane (HPLC grade), and polyethylene glycol 600 (PEG 600, synthesis grade) were purchased from Merck (Darmstadt, Germany). Abietic acid (analytical grade) was purchased from Alfa Aesar (Karlsruhe, Germany), and α-pinene and α-methylstyrene (both analytical grade) were purchased from Sigma-Aldrich (Steinheim, Germany). All reagents were used without further purification. Two different PSAs were taken for analysis. The acrylic adhesive consisted of ethylhexyl acrylate, butyl acrylate, and acrylic acid (59/29/12, w/ w) and was dissolved in acetone/petroleum ether (bp 60 95 °C, 30/70, v/v). The synthetic rubber was a commercially available styrene-isoprene-styrene (SIS) block copolymer. Tackifiers were purchased from different suppliers; further details are listed in Table 1. A resin-modified natural rubber adhesive containing two different tackifiers (sample A) was used to scrutinize the developed method. All tackifier resins were dissolved in toluene to produce a 1% solution (w/w). Each solution was measured using a small melting point capillary with a closed end (Dip-it sampler, IonSense, Saugus, MA). This end was dipped into the solution, and the solvent was allowed to evaporate. Subsequently, the material attached to the surface of the glass rod was directly analyzed in the DART gas stream. Reference substances (abietic acid, α-pinene, and α-methylstyrene) were directly analyzed by placing the samples by means of a melting point capillary in the DART source to verify signals in the resin and PSA spectra. PSAs of two different types were prepared: (i) acrylic-based adhesives and (ii) synthetic-rubber-based adhesives. Four tackifiers of different classes were added to each PSA polymer at mass percentages of 0%, 0.01%, 0.1%, 1%, 10%, 20%, 30%, 40%, and 50%. The rosin resin (resin 1), terpene phenolic resin (resin 6), and polyterpene resin (resin 7) were added to both adhesives (acrylic-based and synthetic-rubber-based), whereas the hydrocarbon resin (resin 8) was added only to the synthetic rubber due to hydrophobicity. The tackifiers were dissolved in toluene and diluted several times to obtain various concentrations. For the preparation of samples, 5 mL of the tackifier resin solution was added to 25 mL of a 10% PSA polymer solution in toluene (w/w). The samples were shaken for 15 min, and subsequently 1 mL of each solution was placed on a siliconized paper. After evaporation of the toluene, a melting point capillary was dipped 7324

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry with its closed end into the PSA layer, so that a small amount of polymer was stuck to the tip. The glass rod was then submitted to analysis directly in the DART gas stream. For the analysis of commercially available PSA tapes, the glass rod was also used to pick up some adhesive material. On the one hand, removing the polymer layer from the backing material directly with the glass tip was sometimes easy, while, on the other hand, sampling of some PSAs proved to be quite difficult. These were gently rubbed with a wooden spatula until a small polymer roll was formed. This roll was removed from the backing with either scissors or the melting point capillary itself. Subsequently, the end of the glass rod with the attached PSA sample was held directly in the DART gas stream. Instrumentation. For the MS analyses, the DART ion source (IonSense, Saugus, MA) was operated in the positive ion mode at 250 °C. Helium was used as the ionizing gas at a flow rate of ∼3 L/min. The electrodes were set as follows: needle voltage 3000 V, grid electrode voltage 250 V, and discharge electrode voltage 150 V. The distance between the ceramic cap of the DART ion source and orifice 1 of the mass spectrometer was approximately 10 mm. Detection was performed on a JEOL AccuTOF orthogonal acceleration single-reflectron time-offlight mass spectrometer (JEOL GmbH, Eching, Germany). To achieve high-resolution mass spectra, the mass scale was calibrated using a solution of 0.5% PEG 600 in 1:1 (v/v) methanol/dichloromethane. The mass spectrometer was operated at a resolving power greater than 6000 (fwhm definition) at m/z 609.28 from protonated reserpine. Thus, a mass accuracy of 2 ppm and less was achieved for all diagnostic signals. Analyte ions were identified by elemental composition calculation (monoisotopic masses, JEOL software). To minimize the background noise, the ion guide potential (peaks voltage) was set to 600 V. Other settings of the mass spectrometer were as follows: ring lens voltage 5 V, orifice 1 voltage 20 V, orifice 2 voltage 5 V. To reduce contamination of orifice 1, the temperature was set to 80 °C. All spectra were acquired in a mass range of m/z 30 1000 and with a spectrum recording interval of 0.2 s. These parameters were chosen because they have been found suitable for all tackifier types. For the evaluation and interpretation of the data, all spectra were background-corrected.

’ RESULTS AND DISCUSSION Rosin Resins. The spectrum of a modified rosin resin, hydrogenated and esterified with glycerol, provided several peaks (Figure 1). As can be seen, the [M + H]+ ion of glyceryl abietate at m/z 377.27 in the rosin resin spectrum is of low abundance as well as the [M + H]+ ion of glyceryl tetrahydroabietate at m/z 381.30. The corresponding signal caused by the fragment of [M H2O + H]+ shows a more than 4 times higher intensity (m/z 363.29). At m/z 305.25, the protonated molecular ion of dihydroabietic acid can be seen. The most abundant peak at m/z 259.24 relates to the [C19H30 + H]+-fragment of abietic acid. Signals of di- and triesters are not very intense. The diester of glycerol and tetrahydroabietic acid and dihydroabietic acid is observed at m/z 649.52 ([M H2O + H]+), whereas triesters are almost not detectable, because the instrument settings were not optimized for the analysis of rosin resin triester signals but, instead, for all four classes of tackifiers. The signals in the rosin resin spectrum show that, even if the resin is modified by hydrogenation, abietic acid and its esters and fragments as well as saturated (di- and tetrahydroabietic acid)

ARTICLE

Figure 1. DART spectrum of resin 1, showing prominent peaks at m/z 259.24, 305.25, 363.29, 377.27, and 649.52. The signals (in the same order) were interpreted as follows: [C19H30 + H]+ fragment of abietic acid, [M + H]+ of dihydroabietic acid, [M H2O + H]+ of glyceryl tetrahydroabietate, [M + H]+ of glyceryl abietate, and [M H2O + H]+ of the diester of glycerol with tetrahydroabietic acid and dihydroabietic acid. At m/z 303.23, the [M + H]+ ion of abietic acid is present (not assigned, see structural formula). The same signal was obtained by analysis of neat abietic acid.

and unsaturated (dehydroabietic acid) derivatives can be detected. For the identification of a rosin resin in an unknown sample, it is, thus, important that signals of a whole peak group are present. The peak groups around the signal of protonated abietic acid at m/z 303.23 (from 301.22 to 307.26) and around the signal of a fragment of abietic acid at m/z 259.24 (from 255.21 to 261.26) are typical for a rosin resin and were found to be present in all spectra of rosin resins which were measured. Nevertheless, spectra of hydrogenated and nonhydrogenated rosin resins can be distinguished. Spectra of three different rosin resins are shown in Figure 2a c. The most intense signal in the peak group around m/z 303.23 in a nonhydrogenated resin is observed at m/z 301.22 corresponding to dehydroabietic acid (Figure 2a), whereas the most abundant peak in a hydrogenated tackifier is observed at m/z 305.25, which relates to dihydroabietic acid (Figure 2b). Tetrahydroabietic acid is also present as [M + H]+ at m/z 307.26 in a hydrogenated rosin resin, while it is almost not detectable in the nonhydrogenated resin. In a disproportionated resin, the signal intensity of abietic acid at m/z 303.23 is rather low, while the intensities of both dehydroabietic acid at m/z 301.22 and dihydroabietic acid at m/z 305.25 are quite high (Figure 2c). By means of these three reproducible patterns, the identification of hydrogenated, nonhydrogenated, and disproportionated rosin resins can be achieved. It is also possible to differentiate between different esters, e.g., a glycerol ester and a pentaerythritol ester. As explained above, glyceryl abietate (hydrogenated) and the corresponding [M H2O + H]+ fragment as well as the glyceryl diabietate signals ([M H2O + H]+, also hydrogenated) were clearly seen in glycerol-modified rosin (Figure 1). In contrast, pentaerythritol-modified rosin yielded protonated molecules of pentaerythrityl abietate (hydrogenated) at the peak groups around m/z 425.33 and around m/z 407.32 for [M H2O + H]+ ions (Figure 3). Diesters were also detectable around m/z 713.57 as protonated molecular ions. Through detection of these characteristic signals, modified rosin esters can be distinguished. 7325

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry

ARTICLE

Figure 3. Mass spectrum of resin 5. Peaks at m/z 259.24 ([C19H30 + H]+ fragment of abietic acid), 305.25 ([M + H]+ of dihydroabietic acid), 407.32 ([M H2O + H]+ of pentaerythrityl tetrahydroabietate), 425.33 ([M + H]+ of pentaerythrityl tetrahydroabietate), and 713.57 ([M + H]+ of diester of pentaerythritol and tetrahydroabietic acid) are present.

Figure 2. (a) Spectrum of a nonhydrogenated rosin resin (resin 2) showing the protonated ion of dehydroabietic acid (m/z 301.22) as the most abundant signal. Abietic acid ([M + H]+ at m/z 303.23) and a resin acid with the molecular formula C20H26O2 ([M + H]+ at m/z 299.20) are also present. (b) DART spectrum of a hydrogenated rosin resin (resin 3) showing an increase in the relative abundance of hydrogenated resin acids (dihydroabietic and tetrahydroabietic acid at m/z 305.25 and 307.26, [M + H]+). (c) A disproportionated rosin resin (resin 4) provides a DART spectrum with a low-abundant abietic acid signal and high-intensity peaks of dehydroabietic and dihydroabietic acid.

Terpene Phenolic Resins. Analysis of a terpene phenolic resin resulted in a number of peaks, as shown in Figure 4a. The peak assigned as 1 is observed at m/z 137.13 and relates to a protonated monoterpene, which was verified by comparative analysis of α-pinene. Although the main compound in this tackifier should be α-pinene, it is not assured that the peak definitely represents α-pinene, due to the huge variability of monoterpene isomers and the fact that they show the same masses. Therefore, the expression “pinene” will be used in this paper for all components with the same molecular formula as α-pinene. Two other signals (2 and 3) represent oligomerized pinene: the protonated dimer and trimer. In addition to the signals which consist only of monoterpene units, there are several

peaks corresponding to reaction products of a monoterpene or monoterpene oligomers with phenol. The most abundant peak in the spectrum at m/z 367.30 represents a compound comprising two pinene units and one phenol unit (assigned as 2 + 1). Other peaks assigned as 3 + 1, 4 + 2, and 5 + 2, corresponding to three pinene and one phenol unit, four pinene and two phenol units, and five pinene and two phenol units, could also be detected. Polyterpene Resins. The spectrum obtained by analysis of a polyterpene resin provided several signals corresponding to different mono-, di-, and oligomers of pinene (Figure 4b). The protonated pinene monomer ion can be seen at m/z 137.13 (peak 1). The base peak corresponds to [2M + H]+ at m/z 273.26 (peak 2). With decreasing abundance, the peaks assigned as 3, 4, 5, and 6 can be observed, which relate to protonated pinene oligomers (trimer, tetramer, pentamer, and hexamer). Oligomers with higher masses are rarely observed under these conditions. As with terpene phenolic resins, it is impossible to distinguish between monoterpene isomers. In all mass spectra obtained for polyterpene resins using the DART ion source, more highly unsaturated species than pinene were also observed, due to the manufacturing process and the complexity of raw materials for production. Moreover, besides the protonated molecules, some M+• ions were detected. Therefore, it is important to note that with regard to rosin resins, the determination of polyterpene resins has to be conducted by means of peak groups and not only single signals, as the polyterpene resins slightly differ from each other. Hydrocarbon Resins. The spectrum in Figure 4c was obtained from a hydrocarbon resin. This resin consists mainly of polymerized hydrogenated dicyclopentadiene (DCPD). As can be seen, analysis revealed a distribution of molecules with several units of DCPD (or its saturated derivatives) and partially cyclopentadiene (or its saturated derivatives). A more detailed interpretation of the spectrum shows that the peaks assigned as 1 4 correspond to the protonated molecules of one to four units of DCPD with different degrees of hydrogenation. Peak 1 relates to monohydrogenated DCPD, whereas peaks 2 4 correspond to 2, 3, and 4 units of oligomerized DCPD with a 2-fold hydrogenation for the whole molecule. The other peaks labeled 1.5 3.5 were identified to represent substances consisting of both hydrogenated DCPD and cyclopentadiene. The 1.5 peak is referred to as [M + H]+ of a monohydrogenated compound consisting of one unit of DCPD and one unit of cyclopentadiene. 7326

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry

ARTICLE

Figure 5. (a) Spectrum of resin 9 showing a complex distribution of hydrocarbon signals. (b) The DART spectrum of resin 10 provided an abundant peak of protonated α-methylstyrene at m/z 119.09 (analysis of neat α-methylstyrene resulted in the same peak). Other signals were not identified.

Figure 4. (a) Spectrum of resin 6, showing protonated ions of a pinene monomer (1), dimer (2), and trimer (3) and protonated ions of combinations of pinene and phenol units (2 + 1, 3 + 1, 4 + 2, and 5 + 2). (b) DART-MS spectrum of resin 7, showing protonated ions of pinene monomer (1, see structural formula) and oligomers from dimer to hexamer (2 6). (c) DART spectrum of resin 8. Peaks of protonated DCPD monomer (see structural formula) to tetramer are present (1 4) as well as protonated molecules consisting of different units of DCPD and cyclopentadiene (varying degrees of hydrogenation; 1.5 3.5). A commonly used stabilizer is assigned by an asterisk.

The two peaks at m/z 335.27 and m/z 467.37 (marked 2.5 and 3.5) correspond to one unit of cyclopentadiene each and two units and three units of DCPD, respectively. Both molecules are 2-fold hydrogenated. Moreover, the spectrum of the commercial product shows a rather weak signal caused by tris(2,4-di(tert)butylphenyl)phosphate (indicated by an asterisk), possibly added as a stabilizer. Because of the different hydrogenation degree, the mass spectra obtained for this tackifier are complex, containing peak groups rather than single signals.

In contrast to the other tackifier resin classes, which are generally composed of one basic component (rosin resins of abietic acid, terpene phenolic resins of pinene and phenol, polyterpene resins of pinene), hydrocarbon resins provide a huge variability in terms of composition. Besides DCPD resins, aliphatic and aromatic resins and monomer resins such as styrene or α-methylstyrene resins are also used. All types of tackifiers can be modified as well. Therefore, the spectra obtained for hydrocarbon resins using DART-MS can be very different, as will be explained in the following examples. A spectrum of a DCPD resin, which shows a clear and defined distribution of polymerized components, was already discussed above. Figure 5a,b shows two spectra of other hydrocarbon resins. The first spectrum which was obtained from an aliphatic hydrocarbon resin consisting of mostly C5 units is very complex, including many signals that are difficult to assign and, therefore, cannot be used as diagnostic signals in an unknown sample. In the second spectrum a poly(α-methylstyrene) resin can be seen. The base peak at m/z 119.09 corresponds to protonated α-methylstyrene (see structural formula). This finding was verified by analysis of neat α-methylstyrene. Dimers or trimers were not detectable in the tackifier resin. The spectrum revealed only a few other signals with rather low intensities which have not been identified. Consequently, the determination of these types of tackifiers in unknown samples is, thus, complicated because of insufficient and nonidentifiable signals. All tackifier resin spectra were thoroughly interpreted and stored in a database to develop a library of DART-MS spectra. Analysis of Tackifier Resins in PSAs. The capability of the DART-MS to ionize tackifier resins of all four classes in different matrixes was investigated as well as the limits of detection (LODs) and the possibility of a semiquantitative analysis by measuring a matrix-based calibration curve. The results are explained and discussed below. 7327

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry

ARTICLE

Table 2. Molecular Formula, Observed and Calculated m/z Ratio, and Mass Error of Characteristic Protonated Ions of Four Selected Resins resin

molecular

observed m/z

formula

([M + H]+)

a

([M + H]+)

(ppm)

rosin resin

C19H30

259.2424

259.2426

0.80

(resin 1)

C20H32O2a

305.2476a

305.2481

1.65

C23H38O3a

363.2896a

363.2899

0.74

C23H36O4 C23H40O4

377.2764 381.3060

377.2744 381.3005

5.32 14.41

C43H68O4

649.5216

649.5196

3.09

137.1331a

137.1330

0.71

terpene phenolic C10H16a C20H32

273.2578

273.2582

1.42

C26H38Oa

367.2996a

367.3001

1.30

C30H48

409.3825

409.3834

2.27

C36H54Oa

503.4248a

503.4253

0.91

C52H76O2 C62H92O2

733.5898 869.6788

733.5924 869.7176

3.54 44.64

polyterpene

C10H16

137.1326

137.1330

3.28

resin (resin 7)

C20H32a

273.2577a

273.2582

1.87

a

a

409.3834

409.3834

0.08

C40H64a

545.5090a

545.5086

0.65

C50H80

681.66365

681.6338

3.94

C60H96

817.7355

817.7590

28.75

C10H14 C15H20

135.1150 201.1639

135.1174 201.1643

17.69 2.04

C20H28

269.2263

269.2269

2.37

C25H34a

335.2738a

335.2739

0.34

C30H40a

401.3205a

401.3208

0.86

a

467.3679a

467.3678

0.29

533.4161

533.4147

2.59

resin (resin 6)

C30H48

hydrocarbon resin (resin 8)

C35H46 C40H52 a

a

calculated m/z m/z error

Three diagnostic signals for each resin were used for detection.

All tackifiers were detectable by the formation of at least three diagnostic signals in both polymers (Table 2). Figure 6a shows a DART spectrum of an acrylic-based adhesive, modified with 1% (w/w) of terpene phenolic resin. The peaks indicated by asterisks represent the signals of the resin. The sample provided the protonated molecular ion at m/z 367.30 as the base peak. The protonated monomer (m/z 137.13) and dimer (m/z 273.26) of pinene as well as a compound consisting of three pinene units and one phenol unit (m/z 503.43) were also present. Other characteristic signals (assigned as 3, 4 + 2, and 5 + 2 in Figure 4a) were also detectable but with very low intensity. As can be seen, the signals of the tackifier can clearly be distinguished from the matrix peaks. Even though the spectrum of the pure acrylic polymer is very complex, all three tackifier resins tested in this experiment (terpene phenolic, polyterpene, and rosin resin, amounts of 1% w/w each) were detected immediately using diagnostic signals given in Table 2. Synthetic rubber and four tackifiers (the three resins mentioned above and a hydrocarbon resin, amounts of 1% w/w each) were selected to conduct further experiments. The samples were measured and analyzed for the presence of tackifier resins. All four tackifiers were successfully detected along with the signals of the polymer. The determination of tackifier resins in synthetic rubber was slightly better, as the intensity of the matrix signals was lower compared to the acrylic polymer.

Figure 6. (a) Spectrum of 1% (w/w) of resin 6 in acrylic adhesive. Peaks of the resin are indicated by asterisks. (b) DART-MS spectrum of sample A containing a rosin resin (indicated by asterisks) and a terpene phenolic resin (indicated by dots).

Concerning DART measurements, the abundance is strongly dependent on the structure of the analyte. Obviously the DART ionization of acrylates which contain oxygen achieves signals that are more intense than signals of synthetic rubber, which consists mainly of hydrocarbons. A mass spectral database search was conducted for every spectrum of the modified PSAs (with 1% w/w of tackifier) to ensure that this method can be used to analyze tackifier resins in unknown samples. All tackifiers were identified correctly. As every analyte was determined by at least three characteristic peaks, the attribution can be regarded as reliable. To investigate the possibility of semiquantitative analyses, tackifier resins were added to two adhesive polymers in various concentrations. It was ensured that the pure PSAs did not show any signals of the tackifier resins. As described above, all analytes were detectable at amounts of 1% (w/w) along with the matrix. Surprisingly, the samples containing 10% (w/w) of tackifier resin did not show any signals of the polymer; the tackifier resin completely dominated the spectrum. With up to 50% (w/w) of tackifier added, no significant difference between all spectra could be observed. The intensity of the diagnostic fragments of each analyte remained approximately the same between 10% and 50% (w/w) of the tackifier resin. Obviously there are huge differences with regard to the ionization between tackifiers and polymers, which presumably occur due to higher volatility or structural features of the tackifier resins. Apparently, the result is an ionization discrimination of the PSA matrix. Semiquantitative analyses of the tackifiers studied in this work are, thus, not possible as there are no sufficient differences and hence no linear correlation in the intensity of analytes in matrix calibration standards between 10% and 50% (w/w). It would certainly be feasible to conduct further experiments to examine the range between 1% and 10% (w/w) of tackifier resin more closely. Nevertheless, the concentration of tackifiers used in 7328

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry commercially available products often is above 10%, which implies that the relevant range would not be covered. Huge differences between the acrylic or synthetic rubber matrix were not observed, except a slightly lower abundance of the synthetic rubber polymer. At a concentration of 10% (w/w) of tackifier resin, both polymers did not show an intense spectrum while, at a tackifier resin concentration of 1% (w/w) and less, the matrix signals were quite prominent. Moreover, the behavior of the different types of tackifiers was relatively similar, as the intensity of tackifier resin in comparison to the matrix was equal for the rosin resin, terpene phenolic resin, and polyterpene resin. Only the hydrocarbon resin produced signals of lower abundances than the other resins. For the evaluation of the DART method, another experiment was carried out. A sample (sample A) containing two different tackifiers (15% of terpene phenolic resin and 15.1% of rosin resin, w/w) was analyzed as described above. The spectrum can be seen in Figure 6b: signals of the rosin resin are indicated by asterisks and signals of the terpene phenolic resin by dots. A DART-MS spectrum of the sample revealed the diagnostic signals of both tackifiers. It can be concluded that the developed method is suitable for the detection of two different tackifier resins in one sample. Other experiments showed that even three different tackifiers (rosin resin besides terpene phenolic and polyterpene resin) in one sample could be simultaneously assigned. With the use of this method, a lot of known samples were analyzed to verify the analytical procedure, followed by analysis of unknown samples, which provided conclusive results. Limits of Detection. To determine the LODs for all tackifiers in both matrixes, samples containing different amounts of tackifier resins were prepared and analyzed by DART-MS. The matrixes per se did not show any tackifier peaks. Spectra showing less abundant signals of the analyte were used to derive LODs. Criteria for the detectability were at least three peaks of protonated molecules of the tackifier resin and a s/n ratio for each specific ion of the background-corrected mass spectrum of >3. All tackifiers were detected in samples containing 0.1% (w/w) of tackifier resin, whereas samples containing 0.01% of analyte did not show any resin signals. While the calculated s/n ratios of the hydrocarbon resin peaks in synthetic rubber were relatively small and close to the value of 3, the ratios of the most abundant signals of polyterpene and terpene phenolic resin in synthetic rubber were about 10-fold higher. Thus, the LOD of these two tackifiers might be even lower in synthetic rubber.

’ CONCLUSIONS The DART-MS technique proved to be a powerful tool for the analysis of tackifier resins in PSAs. It is characterized by high sensitivity and particularly simple handling. Without time-consuming sample preparation and without extraction procedures, the analysis only takes a few minutes and is much faster than other methods, e.g., MALDI-TOF-MS.10 Generated data provide an overview of a large range of analytes. With the use of a database, the interpretation of the obtained spectra is relatively easy and rapid. This method offers the possibility of analyzing tackifiers in PSAs which are analytically difficult to access. The LODs of all four tackifier resins tested in these experiments were e0.1% (w/w). However, the absence of relatable signals in some hydrocarbon resins limits the potential of DART-MS for the analysis of this type of tackifier.

ARTICLE

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Fax: +49 40 4909 182298.

’ ACKNOWLEDGMENT We gratefully acknowledge support by the tesa SE and especially Mr. Nicolai B€ohm and Mr. Ralf Ihrig for providing the samples, raw materials, and useful information. ’ REFERENCES (1) Petrie, E. M. Handbook of Adhesives and Sealants, 2nd ed.; McGraw-Hill: New York, 2007. (2) Sadhra, S.; Gray, C. N.; Foulds, I. S. J. Chromatogr., B: Biomed. Sci. Appl. 1997, 700, 101–110. (3) Ash, M.; Ash, I. Handbook of Fillers, Extenders, and Diluents, 2nd ed.; Synapse Information Resources, Inc.: Endicott, NY, 2007. (4) Bart, J. C. J. Plastics Additives; IOS Press: Amsterdam, The Netherlands, 2006. (5) Nordmark, U. J. Anal. Appl. Pyrolysis 2000, 55, 93–103. (6) Findeisen, A.; Kolivoska, V.; Kaml, I.; Baatz, W.; Kenndler, E. J. Chromatogr., A 2007, 1157, 454–461. (7) Canellas, E.; Nerin, C.; Moore, R.; Silcock, P. Anal. Chim. Acta 2010, 666, 62–69. (8) Brites, M. J.; Guerreiro, A.; Gigante, B.; Marcelo-Curto, M. J. J. Chromatogr., A 1993, 641, 199–202. (9) Wright, S. J.; Dale, M. J.; Langridge-Smith, P. R. R.; Zhan, Q.; Zenobi, R. Anal. Chem. 1996, 68, 3585–3594. (10) Kumooka, Y. Forensic Sci. Int. 2008, 176, 111–120. (11) Cody, R. B.; Laramee, J. A.; Durst, H. D. Anal. Chem. 2005, 77, 2297–2302. (12) Jagerdeo, E.; Abdel-Rehim, M. J. Am. Soc. Mass Spectrom. 2009, 20, 891–899. (13) Ropero-Miller, J. D.; Stout, P. R.; Bynum, N. D. Microgram J. 2007, 5, 34–40. (14) Steiner, R. R.; Larson, R. L. J. Forensic Sci. 2009, 54, 617–622. (15) Fernandez, F. M.; Cody, R. B.; Green, M.; Hampton, C.; McGready, R.; Sengaloundeth, S.; White, N.; Newton, P. ChemMedChem 2006, 1, 702–705. (16) Fernandez, F. M.; Green, M. D.; Newton, P. N. Ind. Eng. Chem. Res. 2007, 47, 585–590. (17) Newton, P. N.; Fernandez, F. M.; Plancon, A.; Mildenhall, D. C.; Green, M. D.; Ziyong, L.; Christophel, E. M.; Phanouvong, S.; Howells, S.; McIntosh, E.; Laurin, P.; Blum, N.; Hampton, C. Y.; Faure, K.; Nyadong, L.; Soong, C. W.; Santoso, B.; Zhiguang, W.; Newton, J.; Palmer, K. PLoS Med. 2008, 5, 209–219. (18) Newton, P. N.; Hampton, C. Y.; Alter-Hall, K.; Teerwarakulpana, T.; Prakongpan, S.; Ruangveerayuth, R.; White, N. J.; Day, N. P. J.; Tudino, M. B.; Mancuso, N.; Fernandez, F. M. Am. J. Trop. Med. Hyg. 2008, 79, 662–669. (19) Williams, J. P.; Patel, V. J.; Holland, R.; Scrivens, J. H. Rapid Commun. Mass Spectrom. 2006, 20, 1447–1456. (20) Yu, S.; Crawford, E.; Tice, J.; Musselman, B.; Wu, J. T. Anal. Chem. 2009, 81, 193–202. (21) Petucci, C.; Diffendal, J.; Kaufman, D.; Mekonnen, B.; Terefenko, G.; Musselman, B. Anal. Chem. 2007, 79, 5064–5070. (22) Jones, R. W.; Cody, R. B.; McClelland, J. F. J. Forensic Sci. 2006, 51, 915–918. (23) Laramee, J. A.; Durst, H. D.; Connell, T. R.; Nilles, J. M. Am. Lab. 2008, 40 (16), 18–20. (24) Nilles, J. M.; Connell, T. R.; Durst, H. D. Anal. Chem. 2009, 81, 6744–6749. (25) Nilles, J. M.; Connell, T. R.; Stokes, S. T.; Dupont Durst, H. Propellants, Explos., Pyrotech. 2010, 35, 446–451. (26) Schurek, J.; Vaclavik, L.; Hooijerink, H. D.; Lacina, O.; Poustka, J.; Sharman, M.; Caldow, M.; Nielen, M. W.; Hajslova, J. Anal. Chem. 2008, 80, 9567–9575. 7329

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330

Analytical Chemistry

ARTICLE

(27) Ackerman, L. K.; Noonan, G. O.; Begley, T. H. Food Addit. Contam., Part A 2009, 26, 1611–1618. (28) Haunschmidt, M.; Klampfl, C. W.; Buchberger, W.; Hertsens, R. Analyst 2010, 135, 80–85. (29) Rothenbacher, T.; Schwack, W. Rapid Commun. Mass Spectrom. 2009, 23, 2829–2835. (30) Rothenbacher, T.; Schwack, W. Rapid Commun. Mass Spectrom. 2010, 24, 21–29.

7330

dx.doi.org/10.1021/ac2011608 |Anal. Chem. 2011, 83, 7323–7330