A poly-(L)-lysine (PLL) film cast on an ATR plate (ATR : attenuated total reflection) and hydrated with D2O (80% relative humidity, 28 degrees C) was exposed to a periodic temperature variation of Delta T/2 = +/-2 degrees C at the mean value T-mean = 28 degrees C. Under these conditions the well known structural change from alpha-helix to beta-sheet occurred reversibly in a time range of minutes. A T-stimulation period of tau(m) = 14.7 min turned out to be adequate to induce significant phase shifts between the boundary conformations alpha-helix/beta-sheet and the conformations of transient species. So far, little experimental information was available on these transients. However, phase-resolved, temperature-modulated excitation infrared (T-MEIR) spectroscopy enabled enhanced experimental separation of strongly overlapped bands. T-MEIR spectra were obtained by phase sensitive detection (PSD) of periodically induced changes in the infrared spectrum of PLL. They feature dynamic changes of the components of the amide I’ and amide II’ bands, which are not easily detectable by means of conventional stationary infrared spectra of PLL because of strong overlapping with other bands. The advantage of MEIR spectra over conventional difference spectra results predominantely from the additional experimental parameter "modulation frequency" omega(m), which enables a spreading of the signal response in a new dimension due to different kinetics exhibited by overlapped band components. The phases and amplitudes of resolved band components were analyzed by means of both curve/phase fitting and cross-correlation analysis, the latter resulting in 2D-FTIR spectra. Both approaches led to consistent results. Cross-correlation analysis, which was based so far on the computation of the synchronous and asynchronous correlation function, was refined by two approaches. Firstly, constraints were used, which insure that the phase lags between two correlated infrared bands are close to a given angle, e.g. 0 degrees (in phase), +/-180 degrees (anti-phase), or +/-90 degrees (out of phase), respectively. Secondly, cross-correlation analysis was also performed with arbitrary phase correlation angles, Both methods enable a quick and rather accurate estimate of the relative phase lag between two modulated absorbances at any wavenumber in the modulation spectrum. Both, the interpretation of the 2D-IR correlation data and of the results from the curve/phase fitting analysis provided new insights in the amide II’ region. Via phase correlation of unknown components in the amide II’ region with assigned absorption bands in the amide I’ region, it was possible to attribute components in the amide II’ region with the same reliability as for amide I’.