NMR Chemical Shifts of Common Laboratory Solvents as Trace Impurities Hugo E. Gottlieb,*Vadim Kotlyar, and Abraham Nudelman* Department of Chemistry, Bar-Ilan University, Ramat-Gan 52900, Israel Received June 27, 1997 In the course of the routine use of NMR as an aid for organicchemistry, a day-to-day problem is the identifica- tion of signals deriving from common contaminants (water, solvents, stabilizers, oils) in less-than-analyti- cally-pure samples. This data may be available in the literature,butthe time involved in searching for it may be considerable. Another issue is the concentration dependence of chemical shifts (especially 1 H); results obtained twoor three decades ago usually refer to much more concentrated samples, and run at lower magnetic fields, than today’s practice. We therefore decided to collect 1 H and 13 C chemical shifts of what are, in our experience, the most popular “extra peaks” in a variety of commonly used NMR solvents, in the hope thatthis will be of assistance to the practicing chemist. Experimental Section NMR spectra were taken in a Bruker DPX-300 instrument (300.1 and 75.5 MHz for 1 H and 13 C, respectively). Unless otherwise indicated, all were run at room temperature (24 ( 1 °C). For theexperiments in the last section of this paper, probe temperatures were measured with a calibrated Eurotherm 840/T digital thermometer, connected to a thermocouple which was introduced into an NMR tube filled with mineral oil to ap- proximately the same level as a typical sample. At each temperature, the D 2O samples were leftto equilibrate for at least 10 min before the data were collected. In order to avoid having toobtain hundreds of spectra, we prepared seven stock solutions containing approximately equal amounts of several of our entries, chosen in such a way as to prevent intermolecular interactions andpossible ambiguities in assignment. Solution 1: acetone, tert-butyl methyl ether, di- methylformamide, ethanol, toluene. Solution 2: benzene, di- methyl sulfoxide, ethyl acetate, methanol. Solution 3: acetic acid, chloroform, diethyl ether, 2-propanol, tetrahydrofuran. Solution 4: acetonitrile, dichloromethane, dioxane, n-hexane, HMPA. Solution 5: 1,2-dichloroethane, ethyl methyl ketone, n-pentane, pyridine. Solution 6: tert-butyl alcohol, BHT, cyclo- hexane, 1,2-dimethoxyethane, nitromethane, silicone grease, triethylamine. Solution 7: diglyme, dimethylacetamide, ethyl- ene glycol, “grease” (engine oil). For D 2O. Solution 1: acetone, tert-butyl methyl ether, dimethylformamide, ethanol,2-propanol. Solution 2: dimethyl sulfoxide, ethyl acetate, ethylene glycol, methanol. Solution 3: acetonitrile, diglyme, dioxane, HMPA, pyridine. Solution 4: 1,2-dimethoxyethane, dimethylacetamide, ethyl methyl ketone, triethylamine. Solution 5: acetic acid, tert- butyl alcohol, diethyl ether, tetrahydrofuran. In D 2O and CD3OD nitromethane was run separately, as the protons exchanged with deuterium in presence of triethylamine. Results Proton Spectra (Table 1). A sample of 0.6 mL of the solvent, containing 1 μL of TMS, 1 was first run on its own. From thisspectrum we determined the chemical shifts of the solvent residual peak 2 and the water peak. It should be noted thatthe latter is quite temperature- dependent (vide infra). Also, any potential hydrogen- bond acceptor will tend to shiftthe water signal down- field; this is particularly true for nonpolar solvents. In contrast, in e.g. DMSO the water is already strongly hydrogen-bonded to the solvent, and solutes have only a negligibleeffect on its chemical shift. This is also true for D 2 O; the chemical shift of the residual HDO is very temperature-dependent (vide infra) but, maybe counter- intuitively, remarkably solute (andpH) independent. We then added 3 μL of one of our stock solutions to the NMR tube. The chemical shifts were read and are presented in Table 1. Except where indicated, the coupling constants, and therefore the peak shapes, are essentially solvent-independent and are presented only once. For D 2 O as a solvent, the accepted reference peak (δ ) 0) is the methyl signal of the sodium salt of 3-(trimeth- ylsilyl)propanesulfonic acid;one crystal of this was added to each NMR tube. This material hasseveral disadvan- tages, however: it is not volatile, so it cannot be readily eliminated if the sample has to be recovered. In addition, unless one purchases it in the relatively expensive deuterated form, it adds three more signals to the spectrum (methylenes 1, 2, and 3 appear at 2.91, 1.76, and 0.63 ppm, respectively). We suggestthatthe re- sidual HDO peak be used as a secondary reference; we find that if theeffects of temperature are taken into account (vide infra), this is very reproducible. For D 2 O, we used a different set of stock solutions, since many of the less polar substrates are not significantly water- soluble (see Table 1). We also ran sodium acetate and sodium formate (chemical shifts: 1.90 and 8.44 ppm, respectively). Carbon Spectra (Table 2). To each tube, 50 μL of the stock solution and 3 μL of TMS 1 were added. The solvent chemical shifts 3 were obtained from the spectra containing the solutes, and the ranges of chemical shifts (1) Forrecommendations on the publication of NMR data, see: IUPACCommission on Molecular Structure and Spectroscopy. Pure Appl. Chem. 1972, 29, 627; 1976, 45, 217. (2) I.e., the signal of the proton for the isotopomer with one less deuterium than the perdeuterated material, e.g., CHCl3 in CDCl3 or C6D5 H in C6D6. Except for CHCl3, the splitting due to JHD is typically observed (to a good approximation, it is 1/6.5 of the value of the corresponding JHH). For CHD2 groups (deuterated acetone, DMSO, acetonitrile), thissignalis a 1:2:3:2:1quintet with a splitting of ca.2 Hz. (3) In contrastto what wassaid innote 2, in the 13 C spectra the solvent signalis due to the perdeuterated isotopomer, and the one- bond couplings to deuterium are always observable (ca. 20-30 Hz). Figure 1. Chemical shift of HDO as a function of tempera- ture. 7512 J. Org. Chem. 1997, 62, 7512-7515 S0022-3263(97)01176-6 CCC: $14.00 © 1997 American ChemicalSociety