Anal. Chem. 1988, 60, 2095-2099 Pilosof, D.; Kim, H. Y.; Vestal, M. L.; Dyckes, D. F. Bionied. Mess Spectrom. 1984, 1 1 , 403-407. Kim, H. Y.; Pilosof, D.; Dyckes, D. F.; Vestal, M. L. J . Am. Chem. SOC. 1984. 106, 7304-7309. Raschdorf, F.; Dahinden, R.: Domon, B.; MuBer, D.; Richter, W. J. I n Mess Spectrometry in the Analysis of Large Molecules; McNeal, C. J., Ed.; Wiley: New York, 1988: pp 49-65. Jones, D. S.; Krolik, S.T. Rapid Commun. Mass Spectrom. 1987, 1 , 67-68. March, J. Advanced Organic Chemistry, 3rd ed.; McGraw-Hill: New York, 1985; Reaction 2-3. Mannlng, J. M. J . Am. Chem. SOC. 1970, 92, 7449-7454.
2095
(24) Frelser, B. S.; Wocdln, R. L.; Beauchamp, J. L. J. Am. Chem. SOC. 1975, 9 7 , 6893-6894. (25) Martinson, D. P., Buttrill, S. E., Jr. Org. Mass Spectrom. 1976, 1 1 , 762-772. (26) Hunt, D-F.; Sethi, S. K. J. Am. Chem. SOC.1980, 102, 6953-6963. (27) Siegel, M. M. Presented at the Texas Symposium on Mass Spectrometry IV: Analysis of Peptides and Profelns; College Station, TX, April 18-20. 1988.
RECEIVED for review February 3, 1988. Accepted June 28, 1988.
Tetrakis(dimethy1amino)ethylene: Identification of Impurities and Compatibility with Common Metal, Polymer, and Ceramic Laboratory Materials Robert T. Rewick* and Mary L. Schumacher S R I International, 333 Ravenswood Avenue, Menlo Park, California 94025 Stephen L. Shapiro and Thomas B. Weber Stanford Linear Accelerator Center, P.O. Box 4349, Stanford University, Stanford, California 94305 Matteo Cavalli-Sforza Santa Cruz Institute for Particle Physics, University of California, Santa Cruz, California 94064
Analytlcal procedures were developed that use gas chromatography and gas chromatographylmass spectrometry techrrlques to determlne the purlty of commercially available and purlfled llquld and gaseous tetrakls(dlmethylamlno)ethylene (TMAE). More than 20 components were detected; most were Identified from their mass spectral fragmentatlon patterns. The six major TMAE lmpurltles were dimethylamine, tetramethylhydrazlne, bls(dlmethylamlno)methane, dlmethylformamlde, tetramethylurea, and tetramethyloxamlde. The major lmpurltles accounted for greater than 99.5 area % of all Impurity components detected. Electron capture analysls of the major TMAE lmpurltles suggested that tetramethyloxamlde and tetramethylurea have relatlve high electron capture cross sectlons campared wlth oxygen. Llquld TMAE was observed to be generally compatible (less than 1% decomposltlon) with 24 commonly used metal, polymer, and ceramlc laboratory materials. I n some cases, however, low concentrations of products were generated that had a high affinity for electron capture. Materlals that formed statlstlcaHy slgniflcant amounts of these products were Identified, A technique was developed that uses gas chromatography to measure the major TMAE lmpurltles In a gas flow stream contalnlng TMAE vapor.
Tetrakis(dimethy1amino)ethylene (TMAE), first reported by Pruett et al. ( I ) , is well-known for its strong electron-donor properties ( 2 , 3 )and its ability to chemiluminesce on reaction with oxygen ( I , 4, 5 ) . Kinetic studies have shown that the thermal decomposition (6) and light-producing reactions (7, 8) of TMAE are complex; chemiluminescence requires catalysis by protonic materials (9). Indeed, pure TMAE in the absence of a few parts per million of.protonic activators such as water and alcohols will not react to any appreciable degree
with oxygen (9). In addition, the major oxidation products, tetramethylurea (TMU) and tetramethyloxamide (TMO), act as quenchers for the light emission (4, 9). The following physical properties of TMAE have been reported: melting point, -4 O C (10);boiling point, 177 O C (11); vapor pressure a t 25 "C, 0.74 Torr (11);refractive index a t 20 O C , 1.4817 (12);liquid conductivity, 1.4 X cm-' (12);mass spectrum, major peaks a t m / z of 200, 185, and 85 (13);NMR, a singlet at 2.59 ppm relative to tetramethylsilane (10); and chemiluminescence maximum, 486 nm in alkane solvents (14). The most unusual property of TMAE, however, concerns the electron-rich environment of the TMAE molecule. Excess electrons in TMAE have a mobility of 2.2 cm2/(V s) at 20 "C, which is 100 times higher than the mobility of electrons in other amines (14). Moreover, TMAE is readily ionized, a property that makes it useful as a liquid or gaseous photocathode (1.516). TMAE and a series of other related tetraaminoethylenes have low gas-phase first ionization potentials (Ip)close to that reported for the lithium atom, 5.39 eV (17). For example, Ipreported for TMAE is 5.36 f 0.02 eV (18). Because tetraaminoethylenes have low ionization potentials, high quantum efficiencies for absorption of ultraviolet light and ejection of a photoelectron, and adequate vapor pressures, there is much interest in the use of these materials to detect Cerenkov radiation (11,19). Studies of the electronic properties of TMAE, such as its photoionization cross section (11, 20), combined with its relatively high vapor pressure (11) suggest that TMAE is the i$eal member of the tetraaminoethylene famiiy for use in a Cerenkov radiation detector. For this reason, particle physicists in the United States and Europe are actively pursuing the use of TMAE in such detectors. Because TMAE is readily ionized to radical cations in the presence of trace amounts of oxygen and protonic activators, TMAE is difficult to maintain in the pure state. Indeed, TMAE has been observed to interact with components of a
0003-2700/88/0360-2095$01.50/00 1988 American Chemical Society
2096
ANALYTICAL CHEMISTRY, VOL. 60, NO. 19, OCTOBER 1, 1988
prototype Cerenkov detector system to generate unidentified impurities that strongly absorb the photoelectrons as they drift in the carrier gas (21). For high detector efficiency, TMAE photoejection electrens must have a relatively long lifetime (about 100 ps) in a Cerenkov drift detector. Materials with high electron capture cross sections will severely reduce electron drift lifetimes; for instance, an oxygen concentration of more than 10 ppm in the drift gas produces unacceptably low lifetimes. Because the relative electron affinity of TMAE reaction/decomposition products is not known, it is possible that the presence of these materials contributes to the low electron drift lifetimes occasionally observed in Cerenkov drift detectors. Aside from numerous reports on reactions of TMAE with oxygen (5,9,22,23),water (13), methane (13),and r acceptors, such as pyrene, anthracene, and nitrobenzenes (2), little information is available on interactions of TMAE with laboratory materials frequently used to construct detectors, gas systems, and transfer lines, such as metals, O-rings, and gaskets. In an unpublished report (24), researchers a t Rutherford Appleton Laboratory describe the compatibility of liquid TMAE with elastomeric materials for O-ring seals, but do not discuss reaction products. Analytical procedures to identify TMAE reaction products with these materials are not well documented in the literature. Carpenter and Bens (25) identified oxidation products of TMAE by using gas chromatography/mass spectrometry (GC/MS). From a total of at least 35 components detected by GC, only seven oxidation products were positively identified by MS. However, the analytical techniques used in this work such as details on detectors, sample size, and mass sweep range/rate, are not well described. Urry and Sheeto (26), using nuclear magnetic resonance (NMR) and mass spectrometry, studied the relative amounts of products formed in the autooxidation of TMAE in various solvents. The results, which were surprisingly independent of both the solvent and the temperature, were as follows: 65% tetramethylurea (TMU), 18% tetramethyloxamide (TMO), 12% tetramethylhydrazine (TMH), and 2 % bis(dimethy1amino)methane (BMAM). Waring and Berard (6) described a GC system for the analysis of TMAE thermal decomposition products. The results from the analysis, using a t least three GC columns, a gas sampling valve, and a vacuum system, suggest the major (98%) products to be methane and dimethylamine (DMA). The thermal decomposition reaction of TMAE is significantly less complex than the oxidation reaction, which can occur at room temperature to form a wide variety of products such as DMA, TMH, and TMU. The objective of our study was to develop improved analytical methods to determine TMAE purity and identify products from the interaction of TMAE with typical laboratory materials used to build particle detectors or flow systems. In this work we optimized the GC method for the analysis of TMAE, confirmed the identity of the major impurities by GC/MS, determined by electron capture analysis several impurities that may have sufficient electron affinity (relative to oxygen) to interfere with the detection of photoelectrons produced in a Cerenkov detector, and evaluated the reactivity of liquid TMAE with commonly used metal, plastic, polymeric, and ceramic materials.
EXPERIMENTAL SECTION TMAE from RSA Corp. and Brookhaven National Laboratory was used in the present work. The RSA TMAE was used directly without further purification. The Brookhaven TMAE, also from RSA Corp., was purified at Brookhaven by a treatment described by Holroyd et al. (12): water washing to remove DMA, TMO, and other water-soluble impurities; drying over 5-8, molecular sieves; passing through a column of silica gel, previously activated
at 400 "C, under an atmosphere of nitrogen; distilling several times under vacuum from trap (+50 "C) to trap (-78 "C); and finally storing over NaK alloy for several days. Samples of TMAE from each successive purification step, except for that stored over NaK, were shipped to SRI in wax-sealed bottles for analysis. All TMAE samples were transferred in a drybox to septum bottles containing dried 5-A molecular sieves. Some of the bottles were capped with Mininert septum valves (Supelco, Inc.) and removed from the drybox for withdrawal of vapor and liquid syringe samples for GC and GC/MS analyses. A portion of TMAE from RSA Corp. stored in the drybox, was used for compatibility testing. Metal (316 stainless steel, copper, Monel, aluminum, tin-lead solder, Cu-Be alloy), plastic (Mylar, nylon, polyethylene, Tygon, Kapton), polymeric (Teflon, Kel-F, Kalrez, Viton, silicone rubber, Buna-N rubber), epoxy (G-10, Shell Epon-820, polyimide PC board, Scotch-Weld DP-190), and ceramic (Macor, Micalex, quartz) test specimens were exposed after cleaning (water,acetone, and air-drying for metal and ceramic materials; no pretreatment for plastic and epoxy materials) to 1 cm3 RSA TMAE liquid in small septum bottles (transferredin the drybox). Vapor and liquid syringe aliquots of TMAE were removed from these bottles after 24 h of exposure testing at 23-25 "C to identify TMAE decomposition products that were generated in the presence of the test samples. A Hewlett-Packard HP-5830 GC equipped with a flame ionization detector (FID) was used in this study. Injection port and detection temperatures were maintained at 100 and 225 "C, respectively; the oven temperature was programmed from 50 "C (with a 3-min hold) to 210 "C (with a 9-min hold) at 10 deg/min. The flow rate (He) was 45 cm3/min at 50 "C for all columns studied. The carrier gas was passed through a purifier to remove moisture but not oxygen prior to use. Columns were glass, 8 f t x 2 mm i.d., and packed with (A) 28% Pennwalt 223 4% KOH on 80/100 mesh Gas Chrom R, (B) Carbopack B/4% CW 20 M/0.8% KOH, (C) 10% Apiezon L/2% KOH on 80/lOO mesh Chromosorb W AW, and (D) 20% SE-52 on 100/120 mesh Chromosorb W AW DMCS. For liquid samples, 1pL was injected. Head space samples were 5 pL. For gas stream analysis of TMAE, a Hewlett-Packard HP5890A GC equipped with an FID and a fused silica column (15 m X 0.53 mm i.d.) coated with DB-5 (1.5 pm thickness) was used. Carrier gas flow (He) was 8.5 cm3/min at 38 "C (make-up 16.5 cm3/min);injector and detector temperatures were maintained at 70 and 225 "C, respectively. Temperature programmed runs were 38 "C (3-min hold) to 210 "C at 10 "C/min. Sample size was a 1-cm3Valco sample loop at 30 "C. A Hewlett-Packard HP-5890A GC, equipped with an electron capture detector, was used to determine the relative electron affinity of TMAE impurities. The same fused silica DB-5 capillary column described above was used isothermally (50 "C) for the analysis of BMAM, DMF, and TMA. For TMU and TMO a temperature program rate of 75 OC/min from 100 "C (5-min hold) to 150 "C (1-min hold) was required. Injector and detector temperatures were maintained at 100 and 225 "C, respectively. Sample volumes of 1 p L of liquid (TMU, TMO), 5 p L of vapor (BMAM),and 10 pL of vapor (DMF, TMH) were used. Analysis for TMU required a column carrier gas (N2)flow of 5 cm3/min with 65 cm3/min flow for make-up; for TMO, a column flow of 60 cm3/min with no make-up flow; and, for BMAM, DMF, and TMH, a column flow of 9 cm3/min and a make-up flow of 43 cm3/min. Oxygen will abs.rb TMAE photoejection electrons and reduce the efficiency of a Cerenkov drift detector (27). The ECD response to major TMAE impurities relative to oxygen was measured by using dilute solutions (7-75 ppm) of TMH, BMAM, DMF, TMU, and TMO in high-purity hexane (Baker Resi-Analyzed). DMA was not tested because of its high vapor pressure and expected low electron capture detector (ECD) response. Samples of oxygen (6-25 ppm in helium) were analyzed by the GC from a l-cm3gas sampling loop. The linear detection range of the ECD was determined from measurements of different oxygen and TMAE impurity concentrations. GC/MS analysis of TMAE from RSA Corp. helped to identify impurities. A Ribermag Model R 10-10GC/MS (Nermag, France) was used under the following conditions: 30-m DB-5-coated capillary column programmed from 50 to 300 "C at 10 "C/min;
[
ANALYTICAL CHEMISTRY, VOL. 60, NO. 19, OCTOBER 1, 1988
Table I. Effect of Purification Treatment on TMAE Puritya
Column
n Le Y
a
4
B
12
16
TIME (minl
20
24
28
0
4
8
1
2
16
20
DMA TMH BMAM DMF TMU TMAE TMO unknown
24
TIME ( m i d
Flgure 1. GC chromatograms for TMAE analyzed on two columns: column A, 10% Apiezon L/2% KOH on 80/100mesh gas Chromosorb W AW; column 8, 20% SE-52 on 100/120 mesh Chromosorb W AW DMCS. Key: (1) DMA, (2) TMH, (3) BMAM, (4) DMF, (5) TMU, (6)
TMAE, (7) TMO.
mass spectral sweeps rate, 50-300 amu s-l; ionizing voltage, 70 V; head pressure, 0.3 bar; solvent flushing mode (discriminated against low molecular weight ( m / z 99.5% of all impurities detected in the presence of each test specimen. Material compatibility results were compared based on differences in concentration of impurities in TMAE liquid and head space vapor in the presence and absence of a given test specimen. If the concentration of a given impurity in the liquid phase exceeded the mean value of the same impurity
in the blank by one standard deviation (SD), and in the head space vapor, by 2 SD, the compatibility of the test specimen with TMAE was suspect. TMAE liquid and head space vapor analysis blanks, based on five to seven replicate analyses each, gave TMAE purity values of 98.56 f 0.17 area % (Table I) and 80.35 f 11.08 area % , respectively. Head space sampling by syringe is typically less precise than liquid phase results. By use of the arbitrary compatibility rating system described above, the following critical concentrations (area %) are: for liquid TMAE impurities (blank mean + 1SD) DMA (0.15), TMH (0.17), BMAM (0.57), DMF (0.09), TMU (0.79), and TMO (0.18); and head space TMAE impurities (blank mean + 2 SD) T M H (2.94), TMU (10.66), and TMO (1.23). Table I1 lists measured impurity concentrations for test specimens in the presence of TMAE that exceed the critical concentrations calculated above. Materials that do not generate critical levels of impurities in the presence of TMAE are solder, Mylar, DP-190, polyimide, Tygon, and Kapton. Although the results suggest that some test specimens may be incompatible with liquid TMAE (based on our arbitrary rating system), the extent of the decomposition is less than 1%in all cases, based on the concentration of the impurities in the TMAE blank (Table I). Trace amounts of water, surface oxides, or other impurities left on samples after cleaning and before exposure to TMAE may account for some of the high critical concentration values shown in Table 11. TMU, TMH, and TMO are reported to be major products from reaction of TMAE (25). T M H and TMU are formed a t high levels with many of the samples tested. Results in Table I1 should be interpreted with some caution because (1)the overall decomposition of TMAE is
