Optimized Quantitative DEPT and Quantitative POMMIE Experiments

Oct 9, 2008 - To overcome these problems, Henderson proposed a quantitative DEPT ... QuantitativeC Traces of Glucose Fate in Hepatitis B Virus-Infecte...
3 downloads 0 Views 328KB Size
Download PDF
Anal. Chem. 2008, 80, 8293–8298

Optimized Quantitative DEPT and Quantitative POMMIE Experiments for 13C NMR Bin Jiang,† Nan Xiao,† Huili Liu,† Zhiming Zhou,†,‡ Xi-an Mao,†,§ and Maili Liu*,† State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, The Chinese Academy of Sciences, Wuhan 430071, China, and Department of Physiology and Biophysics, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106 In quantitative analysis, inverse gated 1H decoupled 13C NMR provides higher resolution than 1H NMR. However, due to the lower sensitivity and longer relaxation time, 13 C NMR experiment takes much longer time to obtain a spectrum with adequate signal-to-noise ratio. The sensitivity can be enhanced with DEPT and INEPT approaches by transferring polarization from 1H (I) to 13C (S), but since the enhancements depend on coupling constants (1JSI) and spin systems (SI, SI2, SI3), the enhancements for different spin systems are not uniform and quantitative analyses are seriously affected. To overcome these problems, Henderson proposed a quantitative DEPT (QDEPT) method by cycling selected read pulse angles and polarization-transfer delays (Henderson, T. J. J. Am. Chem. Soc. 2004, 126, 3682-3683), and satisfactory results for SI system are achieved. However, the optimization is incomplete for the SI2 and SI3 systems. Here, we present an improved version of Q-DEPT (Q-DEPT+) and a quantitative POMMIE (Q-POMMIE) where the cyclic delays and read pulse phases are applied. The improved methods prove to be suitable for all spin systems over a large J-coupling range (90-230 Hz), and the 13C signals are nearly equally enhanced with standard deviation less than 5%. NMR spectroscopy is a powerful tool to provide quantitative information for components in a mixture.1 However, 1H NMR peaks split because of homonuclear coupling and the line widths increase with molecular weight, which often make signals to overlap. Compared with 1H NMR, 13C spectroscopy offers much higher resolution due to its wider chemical shift range and singlet line shape after heteronuclear decoupling. Unfortunately, it takes much longer time to obtain a 13C spectrum with an adequate signal-to-noise ratio (SNR) due to the low sensitivity, long relaxation time, and natural abundance of 13C nuclei. In addition to using high-field magnets and cryogenic probes, there are two ways to enhance 13C sensitivity significantly. Irradiation on 1H nuclei before read pulse can enhance 13C * To whom correspondence should be addressed. Fax: +86 27 8719 9291. E-mail: [email protected]. † Wuhan Institute of Physics and Mathematics. ‡ Present address: Department of Biological Sciences, 14 Science Drive 4, National University of Singapore, Singapore 117543. § Case Western Reserve University. (1) Shoolery, J.; Smithson, L. J. Am. Oil Chem. Soc. 1970, 47, 153–157. 10.1021/ac8015455 CCC: $40.75  2008 American Chemical Society Published on Web 10/09/2008

sensitivity by as much as 294% due to nuclear Overhauser effect (NOE).2,3 However, as the NOE enhancement may not be the same for different 13C nuclei, the nuclear Overhauser enhancement cannot be used in quantitative analysis. The other way is to transfer polarization via scalar coupling from 1H (I) to 13C (S). Because the polarization-transfer efficiency is highly dependent on one-bond coupling constant (1JSI), spin systems (SI, SI2, SI3), and the transfer delay (∆), the sensitivity enhancement is not uniform for 13C spins with different attached protons and 1JSI values, which restricts the application of polarization transfer to quantitative analysis. Sami Heikkinen et al.4,5 developed a quantitative HSQC (Q-HSQC) method with uniform sensitivity enhancement via polarization transfer. In conventional HSQC,6 polarization transfer efficiency is described by eq 1 M ∝ sin2(π∆1JSI)

(1)

where M is signal intensity. In Q-HSQC, uniform signal intensities over an expected range of 1JSI are obtained through averaging HSQC spectra recorded with a set of selected ∆ values. The ∆ values are optimized iteratively by minimizing the difference between maximum and minimum of Σsin2(π∆1JSI) over a 1JSI range. Henderson7 introduced this idea to DEPT8,9 and proposed a quantitative DEPT (Q-DEPT) method for 13C NMR. He optimized the polarization transfer times based on eq 1 and the readpulse flip angles (θ) based on the relation between signal intensity and the read pulse shown in eqs 2-4. MSI ∝ (γI ⁄ γS) sin(θ)

(2)

MSI2 ∝ (γI ⁄ γS) sin(2θ)

(3)

MSI3 ∝ 3(γI ⁄ γS) sin(θ)cos2(θ)

(4)

(2) Overhauser, A. W. Phys. Rev. 1953, 92, 411. (3) Solomon, I. Phys. Rev. 1955, 99, 559. (4) Heikkinen, S.; Toikka, M. M.; Karhunen, P. T.; Kilpelainen, I. A. J. Am. Chem. Soc. 2003, 125, 4362–4367. (5) Koskela, H.; Kilpelainen, I.; Heikkinen, S. J. Magn. Reson. 2005, 174, 237– 244. (6) Bendall, M. R.; Pegg, D. T.; Doddrell, D. M. J. Magn. Reson. 1981, 45, 8–29. (7) Henderson, T. J. J. Am. Chem. Soc. 2004, 126, 3682–3683. (8) Doddrell, D. M.; Pegg, D. T.; Bendall, M. R. J. Magn. Reson. 1982, 48, 323–327. (9) Sorensen, O. W.; Ernst, R. R. J. Magn. Reson. 1983, 51, 477–489.

Analytical Chemistry, Vol. 80, No. 21, November 1, 2008

8293

Figure 1. (A) Q-DEPT+ pulse sequence. The thin and thick black bars are 90° and 180° pulses, respectively. The gray bar is the read pulse (θ). Pulses and receiver are phase cycled as follows: Φ1)2(x), 2(y), 2(-x), 2(-y); Φ2 ) 2(x), 2(y), 2(-x), 2(-y); Φ3 ) y, -y, -x, x, -y, y, x, -x; Φ4 ) 2(x), 2(y), 2(-x), 2(-y); Φ5 ) 2(y), 2(-x), 2(-y), 2(x); Φrev ) x, -x, y,-y, -x, x, -y, y. The polarization transfer delay and the flip-angle of read pulse are also cycled: ∆ ) 8(1.384 ms), 8(1.536 ms), 8(2.173 ms), 8(3.319 ms), 8(3.319 ms), 8(4.234 ms), 8(5.331 ms), 8(7.041 ms). θ ) 64(35.3°), 64(48.0°), 64(50.6°), 64(78.5°), 64(87.5°), 64(87.9°). (B) Q-POMMIE pulse sequence. The thin and thick bars are 90° and 180° pulses, respectively. Pulses and receiver are phase cycled as follows: Φ1 ) x; Φ2 ) x, y; Φ3 ) x; Φ4 ) 2(80.2°), 2(106.3 °), 2(112.4°), 2(114.6°), 2(130.5°), 2(149.7°); Φ5 ) x; Φ6 ) y; Φrev ) x, -x. And the polarization transfer delay cycle is ∆ ) 12(1.331 ms), 12(1.922 ms), 12(1.995 ms), 12(3.321 ms), 12(3.346 ms), 12(4.094 ms), 12(5.292 ms), 12(6.987 ms). Table 1. Optimized Parameters for Q-DEPT and Q-POMMIEa Q-DEPT+ Q-POMMIE

∆ (ms)

θ or φ (°)

1.384, 1.536, 2.173, 3.319, 3.319, 4.234, 5.331, 7.041 1.331, 1.922, 1.995, 3.321, 3.346, 4.094, 5.292, 6.987

35.3, 48.0, 50.6, 78.5, 87.5, 87.9 80.2, 106.3, 112.4, 114.6, 130.5, 149.7

a The numbers of θ (or φ) and ∆ can be chosen arbitrarily in the course of parameter optimization. We found the combination of six θ-values (or φ-values) and eight ∆-values gives good enough curvefitting here.

In Henderson’s version of Q-DEPT, summation of time domain data acquired with the selected ∆ delays (2.67, 3.11, 3.12, and 5.96 ms) can eliminate J dependence described by eq 1 virtually, over the expected 1JSI range. And averaging spectra with suitable θ-values (35.6°, 49.1°, 50.0°, and 84.9°) can give the same sensitivity enhancement for all detectable 13C signals as well. With the uniform polarization-transfer efficiency over an expected 1JSI range, and the same sensitivity enhancements for different 13C nuclei, Q-DEPT can be used for quantitative analysis. By analyzing the polarization-transfer mechanism of Q-DEPT using product operator formalism,10 we noticed that Henderson’s optimization7 was suitable for the SI system but incomplete for the SI2 and SI3 systems. For the 1H-detected Q-HSQC experiment,6 the effect of J-coupling constant on signal intensities for all spin8294

Analytical Chemistry, Vol. 80, No. 21, November 1, 2008

Figure 2. Theoretical sensitivity enhancements in three experiments over the expected 1JCH range (90-230 Hz). The thick lines, thin lines, and dashed lines represent the13C signal enhancement of methyl, methylene, and methine, respectively. (A) The theoretical sensitivity enhancement of Q-DEPT with the original parameters.7 Obviously, the enhancement is not uniform for different types of 13C nuclei. When J-coupling is out of the range, which is from 145 to 190 Hz, the enhancement for CH2 is higher than enhancement for CH by 10% at least. (B) The theoretical sensitivity enhancement of Q-DEPT+ with the parameters in Table 1. (C) The theoretical sensitivity enhancement of Q-POMMIE with the parameters in Table 1.

systems is given in eq 1. However, for the 13C-detected experiment, such as DEPT, there are extra term(s) that must be considered for the SI2 and SI3 systems.8,9 Herein, we optimized the parameters by considering all contributions in the polarizationtransfer scheme and derived a set of parameters with which the sensitivities of all types of protonated 13C nuclei can be enhanced in a similar manner. For convenience, this improved method is called Q-DEPT+. In addition, we noticed that uniform enhancement can be achieved by cycling the read-pulse phases and the transfer delays in POMMIE.11,12 This improved POMMIE is called Q-POMMIE. We demonstrate here that the Q-POMMIE provides similar results as the Q-DEPT+ over a wider range of J-coupling constants. THEORETICAL BASIS Panels A and B in Figure 1 show pulse sequences for Q-DEPT+ and Q-POMMIE, respectively. The major differences between the quantitative versions and the conventional ones7-9,11,12 are that the formers use fixed delays and fixed read-pulse flip angles. Using the product operator formalism,10 one can get full expression of observable terms in the Q-DEPT+ experiment for the three spin systems:8,9

SI : (γI ⁄ γS)Sx sin2(π∆1JSI) sin(θ)

(5)

SI2 : 2(γI ⁄ γS)Sx[sin2(π∆1JSI) sin(θ) cos(θ) + sin2(π∆1JSI) cos2(π∆1JSI) sin(θ)] (6) SI3 : 3(γI ⁄ γS)Sx[sin6(π∆1JSI) sin(θ) cos2(θ) + sin4(π∆1JSI) cos2(π∆1JSI) sin(2θ) + sin2(π∆1JSI) cos4(π∆1JSI) sin(θ)](7) Here protons in the SI2 and SI3 systems are treated equivalently and the homonuclear couplings are ignored. If the delay ∆ is set to a value of (2J)-1, only the terms containing only sinn(π∆1JSI) contribute to the observable signals. When the delay ∆ is offset to (2J)-1, the contribution from the other terms containing cosn(π∆1JSI) will have significant effect on observed signals, which inevitably occurs in Q-DEPT experiments, but these contributions were not considered in the original Q-DEPT optimization.7 To obtain uniform enhancement for different types of 13C nuclei, it is necessary to optimize parameters with consideration of all terms. When θ-values and ∆-values are cycled m and n times, respectively, the signal intensities will be expressed in eqs 8-10. The parameters are optimized by fitting eqs 8-10 to a straight line over a presumed J-coupling range with MLAB13 software. The criteria for the optimization are to maintain possible maximum uniform sensitivity enhancement over a selected J-coupling range. We found that eight ∆-values and six θ-values (Table 1) give rise to sensitivity enhancement of 204% over a J-range of 90-230 Hz for the three spin systems with standard deviation of less 5%. Considering the minimum phase cycling of 2 for the DEPT experiment, the number of scans for the proposed version of Q-DEPT+ is 2 × 8 × 6k (96k, k ) 1, 2, 3, 4...). m

MSI ∝ Sx

n

∑ sin θ ∑ sin (πJ∆ ) 2

j

m

MSI2 ∝ 2Sx

∑

n

sin θj

j)1

∑ [sin (πJ∆ ) cos (πJ∆ )]+ 2

2

i

i

i)1

m

2Sx

∑

n

sin θj cos θj

j)1

∑ sin (πJ∆ ) (9)

2

j)1

4

i

i

i)1

m

3Sx

n

∑ sin2θ ∑ [sin (πJ∆ ) cos (πJ∆ )] + 4

2

j

j)1

i

i

i)1

m

3Sx

n

∑ (sin θ cos θ )∑ sin (πJ∆ ) (10) 2

j

j)1

(11)

SI2 : -2(γI ⁄ γS)Sx[sin4(π∆1JSI) sin(φ) cos(φ) sin2(π∆1JSI) cos2(π∆1JSI) sin(φ)] (12) SI3 : 3(γI ⁄ γS)Sx[sin6(π∆1JSI) sin(φ) cos2(φ) sin4(π∆1JSI) cos2(π∆1JSI) sin(2φ) + sin2(π∆1JSI) cos 4(π∆1JSI) sin(φ)] (13) where φ ()Φ4) is the last 90° 1H-pulse phase (Figure 1B). Using an approach similar to Q-DEPT+, we derived a set of parameters consisting of eight ∆-values and six φ-values for quantitative 13C NMR (Table 1). This set of parameters, theoretically, gives rise to sensitivity enhancement of 208 ± 3% for CH (dash line), 209 ± 5% for CH2 (thin line), and 208 ± 4% for CH3 (thick line), respectively, over a J-coupling range of 90-230 Hz (Figure 2C).

i

i)1

∑ sinθ ∑ [sin (πJ∆ ) cos (πJ∆ )]+ j

SI : (γI ⁄ γS)Sx sin2(π∆1JSI) sin(φ)

4

n

m

MSI3 ∝ 3Sx

(8)

i

i)1

j)1

DEPT+), respectively. It can be seen from the figure that the Q-DEPT gives rise to better sensitivity enhancement, 233 ± 2% for CH (dash line) over J-coupling constant between 115 and 225 Hz. The variations for CH2 (thin line) and CH3 (thick line) were mainly caused by incomplete optimization of J-couplings (Figure 2A). While Q-DEPT+ results in a flat enhancement over J-coupling constants from 90 to 230 Hz for the three types of carbons (Figure 2B): 204 ± 3% for CH (dash line), 205 ± 4% for CH2 (thin line), and 204 ± 5% for CH3 (thick line). The cost here is reduction of ∼12.4% in sensitivities compared with original Q-DEPT.7 Obviously, the quantitative information provided with Q-DEPT+ depends closely with the precision of pulse width. However, in NMR experiments, the precision of pulse width is usually confined in ∼0.1 µs, which corresponds to flip-angle 0.9° when π/2 pulse width is 10 µs. We notice the precision of phase angle of rf pulse can reach to 0.006° in a state-of-the-art NMR instrument.14 So, in addition to Q-DEPT+, we proposed a sensitivity enhancing quantitative method, Q-POMMIE, based on POMMIE.11,12 POMMIE is another method (Figure 1B) for 13C NMR sensitivity enhancement via polarization transfer, where observable signal intensities depend on the transfer delays (∆) and phase (Φ4 ) φ) of the last proton pulse as given in eqs 11-13.

6

j

i

i)1

Panels A and B in Figure 2 show the simulations using Henderson’s parameters7 (Q-DEPT) and the current ones (Q(10) Sorensen, O. W.; Eich, G. W.; Levitt, M. H.; Bodenhausen, G.; Ernst, R. R. Prog. Nucl. Magn. Reson. Spectrosc. 1984, 16, 163–192. (11) Bulsing, J. M.; Brooks, W. M.; Field, J.; Doddrell, D. M. J. Magn. Reson. 1984, 56, 167–173. (12) Bulsing, J. M.; Doddrell, D. M. J. Magn. Reson. 1985, 61, 197–219.

EXPERIMENTAL VALIDATION AND DISCUSSION To test the performances of the Q-DEPT, Q-DEPT+, and Q-POMMIE approaches, we prepared a sample containing 2.9 mmol of ethanol (CH3CH2OH) and 2.76 mmol chloroform (CHCl3) with a molar concentration proportion of ∼100:95, in CDCl3 (0.55 mL). The J-coupling constants are 127 (13CH3), 140 (13CH2), and 209 Hz (13CH). Three repeated measurements were carried out for the single-pulse 1H experiment, inverse gated 1H decoupling 13 C experiment, Q-DEPT, Q-DEPT+, and Q-POMMIE experiments, respectively, on a Bruker Avance2+ 500 spectrometer equipped with a TXI probe head. The relaxation delays were 20 s in 1H experiments and 105 s in 13C experiments, to allow the spin system to return to its full equilibrium state. The widths of π/2 pulses were 17 µs for 13C channel and 10 µs for 1H channel. The receiver gain was set as 45.2 in each of 13C experiments. The number of scans (NS) was 32 for Q-DEPT and 96 for the other 13 C experiments. The spectra were processed conventionally using Analytical Chemistry, Vol. 80, No. 21, November 1, 2008

8295

Table 2. Average Values and Standard Deviations of the Integrals of the 1H the Experimentsa

13

C Signals Obtained from the Five

average integral value with standard deviation CH3 ( JCH ) 127 Hz)

CH2 (1JCH ) 140 Hz)

CH (1JCH ) 209 Hz)

3.0001 ± 0.0001 0.997 ± 0.006 0.997 ± 0.003 0.997 ± 0.003 1.009 ± 0.009

2.0005 ± 0.0008 1.003 ± 0.007 0.898 ± 0.009 1.067 ± 0.002 1.040 ± 0.008

0.9509 ± 0.0001 0.927 ± 0.009 0.740 ± 0.007 1.021 ± 0.009 0.904 ± 0.004

1

method 1

single-pulse H spectrum inverse gated 1H decoupled Q-DEPT Q-DEPT+ Q-POMMIE a

13

C spectrum

The integrals of the methyl peak are calibrated as 3.0000 in 1H spectra, and 1.000 in

Table 3. Sensitivity Enhancing Factors of the Three Methods sensitivity enhancing factor experiment

CH3

CH2

CH

Q-DEPT+ Q-DEPT Q-POMMIE

1.684 2.221 1.746

1.683 2.003 1.683

1.687 1.870 1.638

TopSpin 2.0 (Bruker) with digital resolutions of 0.19 Hz for both time and spectral domain data. One group of 1H spectra and four groups of 13C spectra of ethanol and chloroform were obtained through the experiments above. The integral values of the 1H and 13C signals are listed in

13

C spectra.

Table 2. For convenience of comparison, the integral value of CH3 was calibrated as 3.000 in 1H spectrum and 1.000 in 13C spectra, respectively. Each 13C spectrum shared the same set of processing parameters. And it should be noted that, the methine peak of chloroform overlapped with the triplet of CDCl3, so we got the integral of CH peak through deconvolution. Since the sample was a mixture of ethanol and chloroform with the molar concentration proportion of 100:95, the three integrals of CH3, CH2, and CH should be 3:2:0.95 in 1H spectra and 1:1:0.95 in 13C spectra. From the listed data, it can be seen that the quantitative results from the 1H spectra and inverse gated 1H decoupled 13C spectra accord with the fact very well. As for the three sensitivity enhancing methods, Q-DEPT+ and Q-POMMIE provide similar quantitative information with inverse gated 1H decoupled 13C spectra, while the quantitative information provided by Q-DEPT deviated evidently. It also can be seen that, in Q-DEPT spectra, the integral of CH3 peak of ethanol was the biggest and the integral of CH peak of chloroform was the smallest, which accorded with the prediction of Figure 2A. Table 2 also clearly show that, compared with Q-DEPT+, the integral values from Q-POMMIE were closer to those from inverse gated 1H decoupled 13C spectra, the reason for which may be that the precision of the pulse phase in Q-POMMIE is higher than that of the flip-angle in Q-DEPT+. Apparently, Q-POMMIE can be an alternative 13C quantitative method. Since the SNR is proportional to the square root of scan number, we can calculate the sensitivity enhancing factor with the following equation. En )

Figure 3. Signal intensity ratio (M(S)/M(L)) of two 13C nuclei linked with protons with shorter and longer T2 constant (T2(S) and T2(L)) dependence of T2(S)/T2(L) in (A) Q-DEPT+ and (B) Q-POMMIE experiments. The dashed lines, thin lines, and thick lines show the quantitative results dependence on transverse relaxation when T2(L) is 30, 75, and 200 ms, respectively. 8296

Analytical Chemistry, Vol. 80, No. 21, November 1, 2008

SNR ⁄ √NS SNR0 ⁄ √NS0

(14)

where SNR and NS are the signal-to-noise ratio and scan number in sensitivity enhancing spectrum and SNR0 and NS0 are those in inverse gated 1H decoupled 13C spectrum. Table 3 lists the sensitivity enhancing factor of the three methods. Though the experimental data were generally lower than theoretical prediction, it also can be seen that both Q-DEPT+ and Q-POMMIE provided almost the same sensitivity enhancement to the three 13C peaks, while sensitivity enhancements were much different for the different 13C nuclei in Q-DEPT spectra. Longitudinal and transverse relaxations (T1 and T2) may interfere with quantitative analysis in the Q-DEPT, Q-DEPT+, and Q-POMMIE experiments. When prescan delay is long enough (g5T1), the magnetization is expected to be fully recovered and saturation effect in association with T1 can be avoided. Heikkinen et al.4 had discussed the effect of T2 on the Q-HSQC during the

Table 4. 13C NMR Signal Integrals of 2-Butanol, and the Average Signal-to-Noise Ratios Provided by the Three Quantitative Methodsa integral value experiment inverse gated 1H decoupled Q-DEPT+ Q-POMMIE a

13

C spectrum

CH

CH2

β-CH3

γ-CH3

average SNR

1.000 1.000 1.000

0.962 0.996 1.032

0.963 1.052 1.048

0.933 0.967 0.930

224.76 426.49 430.05

The integral of methine was calibrated as 1.000 in each spectrum.

respectively. The T2 effect on the signal intensities is given in eq 15 for the Q-DEPT+and Q-POMMIE methods. n

M∝

∑ exp(-∆ ⁄ T i

(H) 2 )

exp(-∆i ⁄ T2(CH)) exp(-∆i ⁄ T2(C))

i)1

(15) where ∆i is the polarization-transfer delay, T2(H), T2(C), and T2(CH) represent the transverse relaxation time of single-quantum coherences of 1H and 13C, and multiple-quantum coherence of CH (2IxSy). Assuming 1/T2(CH) ) 1/T2(H) + 1/T2(C) and T2(H)T2(C), eq (15) can be simplified to eq 16. n

M∝

∑ exp(-2∆ ⁄ T i

(H) 2 )

(16)

i)1

13

Figure 4. C spectra of cholesterol in a 0.15 M CDCl3 solution. (A) Inverse gated 1H decoupled 13C spectrum, (B) Q-DEPT+ spectrum, and (C) Q-POMMIE spectrum. The highest peak in (A) is the solution peak of CDCl3. This peak disappears in (B) and (C), because Q-DEPT+ and Q-POMMIE detect the 13C signal transferred from 1H nuclei only.

Figure 5. Integrals of the protonated 13C nuclei given by the inverse gated 1H decoupled 13C spectrum (red), Q-DEPT+ spectrum (blue), and Q-POMMIE spectrum (black). The peak indexes are ranked in descending order of chemical shift. The integral of the 12th peak is ∼2, the reason being that two 13C signals overlap there.

magnetization-transfer periods, where the transverse magnetizations are at single quantum state. In the Q-DEPT+ and QPOMMIE experiments, the quantum states are changes from single-quantum (I), to multiple-quantum (SI) and to singlequantum (S) during the first-, second-, and third-transfer periods,

(13) www.civilized.com. (14) TopSpin pulse programming manual; Bruker BioSpin GmbH.

This is like eq 2 of Q-HSQC,4 where the authors had a detailed discussion for the spins with different T2(H) values. Here, we analyze the T2 effect on signal intensity following the same method. If the T2 values of different 13C-linked protons are of the same order of magnitude, the reliability of the experiment does not compromise much by relaxation effect. However, if the T2 values of these protons are much different, transverse relaxation influences the quantitative information. This influence is illustrated in Figure 3, in which the relative intensities of two signals, M(S)/ M(L), as a function of T2(S)/ T2(L), are presented. Here, the T2(S) and T2(L) are the shorter and longer T2 of 13C-linked protons respectively. Three curves are shown in each panel, where the T2(L) values are 30 (dash line), 74 (thin line), and 200 ms (thick line), respectively. For example, when T2(L) is 200 ms, the difference between the two 13C nuclei signal intensities is smaller than 5%, although T2(S)/T2(L) is 0.6. From Figure 3, it can be concluded that although the relaxation effect increases along with decreasing T2 and T2(S)/T2(L), relaxation has little influence on the quantitative information even in case where T2(L) is short as 30 ms, which is very similar to the case of Q-HSQC.4 Finally, we present two examples to demonstrate the application of these two new quantitative 13C NMR methods. In the first example, the sample was 0.1 mL of 2-butanol (CH3CH2CHOHCH3) dissolved in 0.5 mL of D2O, with 0.07 g/L CuSO4 serving as the relaxation reagent. The inverse gated 1H decoupling 13C, QDEPT+, and Q-POMMIE experiments were performed on the same NMR instrument as before. The relaxation delays were 20 s for all experiments to allow the spin system to return to its full equilibrium state. The number of scans for each experiment was 96. All spectra were processed conventionally with the same set Analytical Chemistry, Vol. 80, No. 21, November 1, 2008

8297

of parameters. From the integral values listed in Table 4, Q-DEPT+ and Q-POMMIE afford quantitative results similar to the conventional 13C NMR experiment, and furthermore, these two methods provide much higher signal-to-noise ratio. As the second example, we show in Figure 4 the application of the new methods to a biomolecule, cholesterol. The inverse gated 1H decoupled 13C, Q-DEPT+ and Q-POMMIE spectra of a 0.15 M CDCl3 solution of cholesterol were collected (Figure 4). All experiments shared the same number of scans, 96, and the same relaxation delay, 65 s, which was long enough to avoid the T1 relaxation effect, and the same set of processing parameters. The spectra in Figure 4 show that the signal-to-noise ratio of Q-DEPT+ and Q-POMMIE spectra are higher than that in the conventional 13C{1H} spectrum. The signal integrals of protonated 13 C nuclei obtained from the three spectra are plotted in Figure 5, in which the satisfactory quantitative results are presented by the new methods again.

enhancement of ∼200% over the J-coupling range of 90-230 Hz and with a standard derivation of less than 5% for the protonated carbons (13CH, 13CH2, 13CH3). Both experimental and theoretical results show the new version of Q-DEPT can reduce the errors for quantification of the 13CH2 and 13CH3 systems, especially when the magnetization-transfer delay (∆) has a larger offset to (21JCH)-1. The interference due to relaxation effect was similar to that in the Q-HSQC. The reliabilities of the proposed Q-DEPT+ and Q-POMMIE were verified experimentally using three different samples.

SUMMARY We presented fully optimized Q-DEPT+ and Q-POMMIE approaches for quantitative 13C NMR with uniformly sensitivity

Received for review July 23, 2008. Accepted September 5, 2008.

8298

Analytical Chemistry, Vol. 80, No. 21, November 1, 2008

ACKNOWLEDGMENT This work is supported by grants from NSFC (20620140104, 20635040), and National Basic Research Program of China (2009CB918603).

AC8015455