1: %\documentclass[]{spie} %>>> use for US letter paper
2: %\documentclass[a4paper]{spie} %>>> use this instead for A4 paper
3: %\documentclass[]{spie}
4: \documentclass[article,aps]{revtex4}
5: \usepackage{graphicx}% Include figure files
6: \usepackage{dcolumn}% Align table columns on decimal point
7: \usepackage{bm}% bold math
8:
9: \addtolength{\voffset}{14mm}
10:
11: \newcommand{\be}{\begin{equation}}
12: \newcommand{\ee}{\end{equation}}
13: \newcommand{\bea}{\vspace{0.25cm}\begin{eqnarray}}
14: \newcommand{\eea}{\end{eqnarray}}
15: \def\NCA{{Nuovo Cimento } A }
16: \def\NIM{{Nucl. Instrum. Methods}}
17: \def\NPA{{Nucl. Phys.} A }
18: \def\PLA{{Phys. Lett.} A }
19: \def\PRL{{Phys. Rev. Lett.} }
20: \def\PRA{{Phys. Rev.} A }
21: \def\PRC{{Phys. Rev.} C }
22: \def\PRD{{Phys. Rev.} D }
23: \def\ZPC{{Z. Phys.} C }
24: \def\ZPA{{Z. Phys.} A }
25: \def\PTP{{Progr. Th. Phys. }}
26: \def\LNC{{Lett. al Nuovo Cimento} }
27:
28: \begin{document}
29: %\usepackage[dvips]{graphicx}
30:
31: \title{ Experimental realization of a measurement conditional
32: unitary operation at single photon level and application to
33: detector characterization. }
34:
35: \author{ M. Genovese}
36: \author{ G. Brida}
37: \author{ M. Gramegna}
38: \author{ M.L. Rastello}
39:
40: \affiliation{ Istituto
41: Elettrotecnico Nazionale Galileo Ferraris, Strada delle Cacce 91,
42: 10135 Torino, Italy}
43:
44: \author{ M. Chekhova}
45: \author{ L. Krivitsky}
46: \author{S. Kulik}
47:
48: \affiliation{ Phys. Dep., M.V. Lomonosov Moscow State Univ.,
49: 119992 Moscow, Russia. }
50: %\authorinfo{ genovese@ien.it. Tel. 39 011 3919253, fax 39 011 3919259, http://www.ien.it/~genovese/marco.html}
51:
52:
53: % \pagestyle{plain}
54: %>>>> uncomment following to start page numbering at 301
55: %\setcounter{page}{301}
56:
57: %\maketitle
58:
59: \begin{abstract}
60:
61:
62: Our last experimental results on the realization of a
63: measurement-conditional unitary operation at single photon level
64: are presented. This gate operates by rotating by $90^o$ the
65: polarization of a photon produced by means of Type-II Parametric
66: Down Conversion conditional to a polarization measurement on the
67: correlated photon. We then propose a new scheme for measuring the
68: quantum efficiency of a single photon detection apparatus by using
69: this set-up. We present experimental results obtained with this
70: scheme compared with {\it traditional} biphoton calibration. Our
71: results show the interesting potentiality of the suggested scheme.
72:
73: \end{abstract}
74: \maketitle
75: \section{Introduction}
76:
77: The possibility of performing a unitary transformation on a qubit
78: conditionally on the result of the measurement on another one is a
79: fundamental tool for Quantum Information \cite{NC}. For example,
80: in teleportation protocols \cite{teleth}, teleportation is
81: obtained by Bob doing an opportune unitary transformation on its
82: sub-part of an entangled system after having received classical
83: information on the result of a Bell measurement performed by Alice
84: on her sub-part of the entangled system and the unknown state to
85: be teleported. Similarly, in entanglement swapping protocols a
86: joint measurement of two members of two distinct entangled pairs
87: allows, by means of conditional unitary transformation, to create
88: a desired entangled state between the two surviving
89: members.
90:
91: Nevertheless, in most of experiments realized up to now with
92: single photons both for teleportation \cite{tele} and swapping
93: \cite{swex}, this part of the
94: protocol was not accomplished, and only correlation measurements with a fixed polarization selection
95: were obtained. The only exception is Ref.~\cite{deMartini}, where a Pockels cell was used
96: for conditional transformations in an active teleportation protocol. Conditional transformations were
97: also performed in experiments where one of the photons of an entangled pair
98: triggered polarization rotation on the other one, to increase the
99: efficiency of the parity
100: check~\cite{pitt1} and to produce signal photons on 'pseudodemand', i.e., in a given time interval
101: after the trigger detection event~\cite{pitt2}.
102:
103: In this paper we present the realization of a set-up addressed to
104: perform conditional unitary transformations on the polarization
105: state of a photon belonging to a polarization-entangled pair.
106: This experiment is the first demonstration of 'purifying' the
107: polarization state of a photon that is initially prepared in a
108: mixed polarization state (initial degree of polarization is equal to zero).
109: As a result of transformations, the degree of polarization increases; its final value is given
110: by the quantum efficiency of the 'trigger' detector. Due to this
111: fact, one can notice another important advantage of our experiment: it can be used for the absolute calibration of
112: photodetectors (absolute measurement of quantum efficiency), a property with large interest for various applications.
113: In fact, in the recent years single photon detectors have found
114: various important scientific and technological applications, which
115: demand for a precise determination of the quantum efficiency of
116: these apparatuses. Among them the studies about foundations of
117: quantum mechanics \cite{au} (and in particular Bell Inequalities
118: measurements \cite{bell,typeI}), quantum cryptography \cite{QCr},
119: quantum computation \cite{QCo,NC}, etc.
120:
121: Classical calibration schemes are based on the use of a strongly
122: attenuated source whose (unattenuated) intensity has been
123: measured by means of a radiometer. The precision of this
124: kind of measurement is limited by the uncertainty in the
125: calibration of the high insertion loss required for reaching
126: single photon level.
127:
128: An alternative is offered by the use of photons produced by means
129: of parametric down conversion (PDC). These states have the
130: property that photons are emitted in pairs strongly correlated in
131: direction, wavelength and polarization. Furthermore, photon of the
132: same pair are emitted within tens of femtoseconds. Since the
133: observation of a photon on a certain direction (signal) implies
134: the presence of another on the conjugated direction (idler), if
135: this last is not observed this depends on the non-ideal quantum
136: efficiency of the detector, which can be measured in this way
137: \cite{bp1,alan}. This method is now approaching a metrological
138: level.
139:
140: A comparison between this last scheme and our new one will be
141: presented in the following.
142:
143: \section{ Description of the measurement-conditioned unitary gate }
144:
145:
146: Our scheme consists in measuring the polarization of a photon
147: belonging to a polarization entangled pair and to perform a
148: unitary operation, by means of a Pockels cell, on the second
149: member of the pair conditional to the result of this measurement.
150:
151: In order to realize this set-up, first of all one needs to produce
152: polarization entangled states of photons by using parametric
153: fluorescence.
154:
155: This state can be generated either by using type-II PDC
156: \cite{bell} or the superposition of two type I PDC emissions
157: \cite{typeI}, obtaining the Bell state:
158:
159: \begin{equation}
160: \vert \psi ^+ \rangle = {\frac{ \vert H \rangle \vert V \rangle +
161: \vert V \rangle \vert H \rangle }{\sqrt {2}}} \label{Psi}
162: \end{equation}
163: where $H,V$ denote horizontal and vertical polarization
164: respectively (other phases between the two components can also be
165: obtained).
166:
167: Then, one of the photons of the pair is addressed, after a
168: polarization selection, to a first detector.
169:
170: When this photon (signal) is detected, we modify the polarization
171: of the delayed second photon (idler) of the pair by means of a
172: Pockels cell driven by a high voltage supply controlled by the
173: output of the first detector.
174:
175: In particular a rotation of $90^o$ of the polarization of the
176: second photon can be produced after the first photon is detected
177: after a specific polarization selection. This realizes the desired
178: measurement-conditional unitary transformation.
179:
180: The result of this unitary operation is that the state of the
181: idler photon, which was initially mixed, becomes partly purified
182: due to this transformation. In terms of polarization properties,
183: light became partly polarized (polarization degree became equal to
184: the quantum efficiency of the trigger detector) while initially it
185: was completely non-polarized (polarization degree was zero).
186:
187: It is worth mentioning that variation of the polarization degree,
188: which we realize in our experiment, is impossible by means of only
189: linear lossless optical methods~\cite{pd}. However, in the present
190: experiment, the transformation performed over the signal photon is
191: essentially nonlinear since it is triggered by the detection of
192: the idler photon entangled to this photon.
193:
194:
195: On the other hand, it must be noticed that in this way the final
196: polarization state of the second photon depends on the quantum
197: efficiency of the first detector, since the Pockels cell is
198: activated only when the first photon is effectively detected. This
199: property, representing a limit for the operation of the
200: measurement-conditional unitary gate, gives nevertheless the
201: opportunity for realising an innovative scheme for absolute
202: calibration of single-photon detectors, which will be discussed
203: below.
204:
205: \section{Experimental realization of the measurement-conditional
206: unitary gate}
207:
208: In our set-up (see Fig. 1,2) we have generated biphoton states by
209: pumping with an argon laser at 351 nm a BBO crystal (5x5x5 mm) cut
210: for producing type II Parametric Down Conversion. Residual pump
211: beam after passing through the non-linear crystal was absorbed by
212: a beam dumper.
213: Then we have selected the directions where
214: the vertically polarized and horizontally polarized emitted
215: circles, corresponding to both photons having a 702 nm wavelength,
216: intersect. A quartz crystal follows the BBO crystal in order to
217: compensate the walk-off of ordinary and extraordinary rays due to
218: birefringence in BBO. In this way indistinguishability between the
219: two polarizations is restored generating the Bell state of Eq.
220: \ref{Psi}.
221:
222:
223:
224: After crossing a polarizing beam splitter (PBS) the first
225: (signal) photon was detected by means of a Perkin-Elmer single
226: photon detector preceded by a pinhole and a red filter. The second
227: photon (idler) was delayed by 200 m of optical fiber. Input and
228: output fiber coupling were realized with 20x objectives with a 0.4
229: numerical aperture. At the output of the fiber this photon was
230: then addressed to a KDP Pockels cell (supplied at 5.2 kV) followed
231: by a Glan-Thompson polarizer, an interference filter at 702 nm (4
232: nm FWHM) and a second Perkin-Elmer Silicon avalanche single photon
233: detector.
234:
235: When the Pockels cell is not active the reduced polarization
236: density matrix of the second photon corresponds to a completely
237: unpolarized case. In fact, the observed intensity of the signal
238: measured when varying the setting of the Glan-Thompson polarizer
239: is flat (see Fig. 3).
240:
241:
242: On the other hand, when the Pockels cell is active the system
243: realizes a rotation of the polarization of the second photon
244: conditioned to a polarization measurement on the conjugated arm.
245: If we choose to perform a rotation by $90^o$ in the same basis of
246: the polarization measurement of the first photon, the result is a
247: purification of the polarization state of the second one.
248:
249: Our experimental results at coincidence level are shown in Fig. 4
250: in the case where we choose the $45^o -135^o$ basis both for
251: polarization measurement and polarization rotation (further
252: results will be presented in a following paragraph when discussing
253: detector calibration). In this basis the state is
254: \begin{equation}
255: \vert \phi ^- \rangle = {\frac{ \vert 45 \rangle \vert 45 \rangle
256: - \vert 135 \rangle \vert 135 \rangle }{\sqrt {2}}} \label{phi}
257: \end{equation}
258: As expected for an entangled state the maximum of coincidences is
259: shifted of $90^o$ when the Pockels cell is active: when the first
260: detector is preceded by a polarizer at $- 45^o$ the maximum of
261: coincidences is when the polarizer before the second detector is
262: set at $135^o$ with the Pockels cell off and at $45^o$ with the
263: Pockels cell on. The visibility is 87.2 \% when the Pockels cell
264: is off and 86.3 \% when it is on.
265:
266:
267:
268:
269: \section{ Description of the method for calibrating single photon detectors.}
270:
271:
272: This scheme for realizing a measurement conditional unitary gate
273: can be applied to the calibration of single photon-detectors. The
274: method is based on the fact that in the real situation the
275: efficiency $\eta_1$ of the first detector is smaller than unity
276: and thus only a fraction of incident photons will be observed and
277: produce an effect on the Pockels cell. Therefore the final density
278: matrix for the second photon does not correspond to the pure
279: state: \be \rho = \vert V \rangle \langle V \vert \ee but to a
280: mixed one:
281: \be
282: \rho =1/2 [ (1+\eta_1) \vert V \rangle \langle V \vert + (1-
283: \eta_1) \vert H \rangle \langle H \vert ] \ee which depends on the
284: quantum efficiency $\eta_1$ of the first detector, allowing a
285: calibration of it. If we insert a polarizer in front of the second
286: detector (Fig.1) and vary its angle $\theta$ (with respect to the
287: horizontal axis), the count rate $N_2$ of the second detector will
288: be given by: \be N_2=N_0 \cdot \alpha \eta_2 [ 1 - \eta_1 cos(2
289: \theta) ] \ee where $\alpha$ represents the idler optical path
290: loss (fiber and Pockels cell transmittances), $\eta_2$ is the
291: quantum efficiency of the second detector and $N_0$ the rate of
292: emission of entangled pairs.
293:
294:
295:
296:
297: The visibility $V$ of the signal counting rate $N_2$ obtained by
298: rotating the polarizer (G) preceding the second detector does not
299: depend on the quantum efficiency $\eta_2$ while it is directly
300: determined by the value of $\eta_1$,
301:
302: \be V = { N^V_2 - N^H_2 \over N^V_2 + N^H_2 } = \eta_1 \ee
303: representing therefore a measurement of this ($N^H_2$ and $N^V_2$
304: are the counts on the second detector when the polarizer is set at
305: $0^o$ and $90^o$ respectively). No coincidence measurement is, in
306: principle, necessary.
307:
308: Incidentally, this effect can be also described in terms of the
309: Stokes parameters, which are initially $S_0=N$, $S_1=S_2=S_3=0$,
310: where $N$ is the photon number, but become, as a result of
311: transformation, $S_0=N$, $S_1= \eta_1 N$, $S_2=S_3=0$. From the
312: Stokes parameters, one can find the polarization degree, which is
313: standardly defined as \be
314: P=\sqrt{\frac{S_1^2+S_2^2+S_3^2}{S_0^2}}.\ee We see that without
315: the transformation, $P=0$ and in the presence of the
316: transformation, $P=\eta_1$.
317:
318:
319:
320: The same effect can also be described in terms of the von Neumann
321: entropy $S$, whose value characterizes the purity of a state,
322: going from zero for a pure state to unity for a completely mixed
323: one. In our case, if $\eta_1=1$ the final polarization reduced
324: density matrix $\rho _2$ of the second photon, corresponding to a
325: pure state, gives $S(\rho_2)=0$. On the other hand if $\eta_1=0$
326: the final polarization reduced density matrix of the second photon
327: is completely mixed with $S(\rho _2)=1$. Intermediate cases lie
328: between these two values (see Fig. 5).
329:
330:
331: \section{ Experimental results for the calibration scheme}
332:
333:
334:
335:
336: The set-up of the calibration scheme is substantially identical to
337: the one presented in section 3, only the idler photon is in this
338: case delayed by 50 m of single mode (4 $\mu m$ core) polarization
339: mantaining fiber. The detection apparatus driving the Pockels cell
340: (including filters and iris) is the device under calibration.
341: Nevertheless, in this case one does not need to have an entangled
342: state and the quartz crystal compensator is not necessary.
343: Therefore, for no compensator introduced the produced state is a
344: mixed one:
345: \begin{equation}
346: \rho = {\frac{ \vert HV\rangle \langle HV \vert + \vert VH \rangle
347: \langle VH \vert }{2}}. \label{rho}
348: \end{equation}
349:
350: Measurements are performed in the $0^o-90^o$ basis.
351:
352: As before, when the Pockels cell in not activated the polarization
353: degree of the second photon is zero. When the Pockels cell is
354: active the system realizes a rotation of $90^o$ on the
355: polarization of the second photon conditioned to a measurement of
356: a vertically polarized photon on the conjugated arm. The results
357: of our measurement under this condition are shown in Fig. 6 and
358: 7. When the cell is active the measured signal has a $(1+ \eta_1
359: cos (2 \theta))$ behaviour as a function of the polarizer setting
360: $\theta$. When the background, estimated by rotating of $90^o$ the
361: pump laser polarization, is subtracted the data show a $(36.8 \pm
362: 2.4) \%$ visibility.
363:
364:
365: This result represents a rough measurement of the quantum
366: efficiency that must be corrected for dead time of the system and
367: the efficiency of the Pockels cell.
368:
369: The Pockels cell driver, when triggered by detector under
370: calibration, generates a high-voltage pulse with fast rising edge
371: (5 ns), a 100 ns flat-top and a long fall tail of about 3.5 $\mu
372: s$ duration (Fig.8). When the mean trigger rate exceeds $10^4$
373: counts per second the Pockels cell driver is disabled for 1 s: in
374: order to make negligible this effect the counting rate on the
375: detector under calibration must be kept lower than the rate
376: threshold of $10$ kHz.
377:
378:
379:
380: For the sake of completeness in Fig.9 we show the single counts on
381: the second detector for horizontal and vertical polarizations in
382: function of the delay and in Fig. 10 the same for coincidences. It
383: is clearly seen as the polarization of second photon is dominantly
384: vertical when the Pockels cell is on, whilst the two polarization
385: are equivalent, at single-count level, when the delay is too
386: large, so that the photon is received before the Pockels cell has
387: been activated. At coincidence level, when the Pockels cell is
388: activated, coincidences are observed for a vertical polarization.
389: When the delay is increased, the situation is reversed and
390: coincidence between vertically (first detector) and horizontally
391: polarized (second detector) photons is observed.
392:
393:
394:
395: The effect of a finite efficiency of the Pockels cell apparatus
396: can be precisely estimated by measuring the visibility at
397: coincidence level (Fig.4). In fact in coincidences the effect of
398: the quantum efficiency of the detector under calibration is
399: irrelevant and the reduction of visibility is completely due to
400: the non-ideal Pockels cell apparatus. From the visibility at
401: coincidence level the efficiency of the Pockels cell apparatus can
402: be estimated to be $0.832 \pm 0.0023$ (see Fig.3). When this
403: further correction is introduced we have \be \eta_1 = { N_V - N_H
404: \over N_V + N_H } \cdot { N^c_V + N^c_H \over N^c_V - N^c_H } \ee
405: where $N_H$ and $N_V$ are the counts on the second detector when
406: the polarizer is set at $0^o$ and $90^o$ respectively and $N^c_H$
407: and $N^c_V$ the corresponding coincidence counts.
408: From our data the measured quantum efficiency of the detection apparatus is
409: then $ 0.441 \pm 0.045$.
410:
411:
412:
413: The uncertainties propagation formula can be written as
414: \be
415: u^2(\eta_1) = c_1^2 u^2(N_H) + c_2^2 u^2(N_V) + c_3^2 u^2(N^c_H) +
416: c_4^2 u^2(N^c_V) \ee
417:
418: The $u$ are the uncertainties and the
419: sensitivity coefficients $c_i$ are evaluated by standard
420: propagation of uncertainties method.
421:
422:
423: A summary of the uncertainty budget is given in Table 1.
424:
425: \begin{table}\begin{center}
426: \begin{tabular}{|l|r|l|r|l|r|} %6 colonne left, right, left
427: \hline
428: \begin{scriptsize}
429: Quantity \end{scriptsize}& \begin{scriptsize} Value \end{scriptsize} & \begin{scriptsize}
430: Standard Deviation \end{scriptsize}& \begin{scriptsize} Type of Distribution \end{scriptsize}&
431: \begin{scriptsize} Sensitivity Coefficient \end{scriptsize}& \begin{scriptsize} Uncertainty contribution
432: \end{scriptsize}
433: \\ \hline
434: $N_H$ &76.6& 4.2& Gaussian & -0.006763 & 0.02840
435: \\ \hline
436: $N_V$ & 165.9 & 5.7 & Gaussian & 0.003123 & 0.01780
437: \\ \hline
438: $N^c_H$ & 4.4 & 1.6 & Gaussian & 0.01827 & 0.02923
439: \\ \hline
440: $N^c_V$ & 48.7 & 2.6 & Gaussian & -0.00165 & 0.00429
441:
442: \end{tabular}
443:
444: \caption{Uncertainty budget for single photon detector calibration
445: with the proposed scheme. Counts are per second.} \label{T1}
446: \end{center}
447: \end{table}
448:
449:
450:
451:
452: Finally, in order to compare with other calibration techniques, it
453: is worth introducing a further correction due to losses in
454: polarizer cube. When this correction is made ($\epsilon = 0.9842$)
455: the final result is $\eta_1 = 0.448 \pm 0.045$ (if the correction
456: for a small drift of the pump power shift is made by taking into
457: account the change in the signal counts, the result becomes $0.454
458: \pm 0.032$ ). Incidentally, even a better precision can be
459: obtained by using a least square fit method applied to the data
460: presented in Fig.6, which leads to $\eta_1 = 0.486 \pm 0.011$
461: \cite{nosCPEM}.
462:
463:
464:
465: For the sake of clarity, it must be emphasized again that the
466: reported quantum efficiency is not the "naked" detector one, but
467: the one corresponding to the detection apparatus including spatial
468: and spectral filtering. In many experimental situations this is
469: the datum necessary for understanding the performances of the
470: set-up. If one wants to measure the "naked" detector quantum
471: efficiency it would be necessary to introduce a corrective factor
472: keeping into account losses in the other elements in the detection
473: apparatus and in the non-linear crystal (corrections are of course
474: needed also for the other calibration method based on PDC). This
475: evaluation is beyond the purposes of this proof-of-principle
476: experiment.
477:
478: Incidentally, for the sake of completeness, we would like to
479: report that similar results ($\eta = 0.417 \pm 0.024$), albeit
480: less accurate (even if a smaller statistical uncertainty) due to a
481: low efficiency of the Pockels cell, were also obtained by using a
482: Lithium Iodate Pockels cell.
483:
484:
485:
486: \section{ Comparison with the traditional calibration
487: scheme with biphotons}
488:
489: In order to check the result obtained with the new scheme, we have
490: compared it with the traditional scheme of single photon detector
491: calibration by using biphotons.
492:
493: In more detail, this procedure \cite{bp1,alan}
494: consists of placing a couple of photon counting detectors
495: down-stream to the non-linear crystal, along the direction of
496: propagation of correlated photon pairs for a selected couple of
497: frequencies: the detection of an event on one detector guarantees
498: with certainty, thanks to the PDC biphotons properties, the
499: presence of a photon on the conjugated direction with a determined
500: wavelength. If N is the total number of photon pairs emitted from
501: the crystal in a given time interval $T_{gate}$ and $N_{signal}$,
502: $N_{idler}$ and $N_{coincidence}$ are the mean number of events
503: recorded, in the same time interval $T_{gate}$, by signal
504: detector, idler detector and in coincidence, respectively, we have
505: the following obvious relationships: \be N_{signal} = \eta_{
506: signal} N = \eta_{ idler} · N \ee where $\eta_{signal}$ and $
507: \eta_{idler}$ are the detection efficiencies on signal and idler
508: arms. The number of events in coincidence is \be N_{coincidence} =
509: \eta_{ signal} \eta_{idler}·N \ee due to the statistical
510: independence of the two detectors. Then the detection efficiency
511: $\eta_{signal}$ follows: \be \eta_{ signal} = N_{coincidence} /
512: N_{idler} \ee
513:
514: This simple relation, slightly modified by taking into account
515: background subtraction and corrections for acquisition apparatuses
516: \cite{bp2}, is the base for the scheme for absolute calibration of
517: single photon detectors by means of PDC light, which reaches now
518: measurement uncertainty competitive with traditional methods
519: \cite{alan,bp2}.
520:
521: When we have applied this scheme to the same configuration as
522: described in the previous paragraph, our result has been $\eta =
523: 0.4812 \pm 0.0015$, which includes the corrections \cite{bp2} for
524: detector dead time ($\tau=40 $ns) $\gamma = 1 - N_{signal} \tau$
525: and for the delay between start and stop signal in Time to
526: Amplitude Converter ($T= 9.3 $ns) $\alpha = 1 - N_{signal} T$. The
527: data from which this result is derived are shown in Fig.6 in
528: function of the counts on the detector under calibration (varied
529: by varying the power of pump laser) both with and without
530: corrections for detector dead time and for stop delay time.
531:
532: For the sake of exemplification the uncertainty budget, for the
533: lower laser intensity, is reported in table 2.
534:
535: This result is perfectly compatible with the one obtained with the
536: new scheme, $\eta_1 = 0.448 \pm 0.045$ (or $\eta_1 = 0.486 \pm
537: 0.011$ with least squares fit method) . At the moment the
538: uncertainty is larger with the proposed scheme, but a substantial
539: reduction of this can be expected by a further careful
540: metrological analysis and determination of all the corrections and
541: an optimization of the Pockels cell apparatus.
542:
543: Thus, these first results show that the proposed method could
544: allow an accuracy comparable with the existing ones and, in
545: particular, could be competitive with the one based on measurement
546: of coincidences. It is therefore worth of further accurate
547: metrological studies.
548:
549: \begin{table}\begin{center}
550: \begin{tabular}{|l|r|l|r|l|r|} %6 colonne left, right, left
551: \hline
552: \begin{scriptsize}
553: Quantity \end{scriptsize}& \begin{scriptsize} Value \end{scriptsize} & \begin{scriptsize}
554: Standard Deviation \end{scriptsize}& \begin{scriptsize} Type of Distribution \end{scriptsize}&
555: \begin{scriptsize} Sensitivity Coefficient \end{scriptsize}& \begin{scriptsize} Uncertainty contribution
556: \end{scriptsize}
557: \\ \hline
558: $N_{i}$ & 1832.8& 9.0& Gaussian & -0.00026 & 0.00234
559: \\ \hline
560: $N_{c}$ & 874.4 & 5.2 & Gaussian & 0.000546&
561: 0.00284
562: \\ \hline
563: $N_{s}$ & 131777 & 185 & Gaussian & $5.88 \cdot 10^{-10}$ &
564: $1.1 \cdot 10^{-7}$
565: \\ \hline
566: T & 9.3 ns & 0.5 ns & Rectangular & 1572 & $7.9 \cdot 10^{-7}$
567:
568: \end{tabular}
569:
570: \caption{Uncertainty budget for single-photon detector calibration
571: with traditional PDC scheme (point at lowest pump power). Counts
572: are per second.} \label{T1}
573: \end{center}
574: \end{table}
575:
576:
577:
578: \section{ Conclusions}
579:
580:
581: In conclusion we have described the realization of a
582: measurement-conditional unitary gate on biphoton states based on
583: the action of a Pockels cell on a member of a polarization
584: entangled pair of photons conditional to a polarization
585: measurement on the other member. This scheme can find various
586: application to Quantum Information processing \cite{NC} and
587: studies of Foundations of Quantum Mechanics \cite{au}.
588:
589: This unitary gate has then been
590: applied for giving a proof of principle of a new method for
591: absolute calibration of detectors based on rotation of
592: polarization of a member on a PDC biphoton state conditioned to
593: detection of the other member after polarization selection, since
594: the polarization degree of the second state is given by the
595: quantum efficiency of the detector driving the Pockels cell.
596:
597: These first results show that the proposed method could allow an
598: accuracy comparable with the existing ones and, in particular,
599: could be competitive with the one based on measurement of
600: coincidences. Of course, a further deep investigation of all the
601: details of this scheme and the use of a system realized on purpose
602: will be necessary for really reaching accuracy levels needed for
603: metrological applications.
604:
605: It must be noticed that in principle this method could be used
606: even if the second detector is an analog one. Furthermore, we are
607: studying the possibility of extending this scheme for calibrating
608: analog detectors as well.
609:
610:
611: \vskip 0.5cm
612:
613: {\bf Acknowledgments}
614:
615: We acknowledge support of INTAS, grant \#01-2122.
616:
617: One of us (L.Krivitsky) acknowledges the support of INTAS YS
618: fellowship grant (03-55-1971).
619:
620: Turin group acknowledges the support of MIUR (FIRB
621: RBAU01L5AZ-002; Cofinanziamento 2001) and Regione Piemonte.
622:
623: Moscow group acknowledges support of the Russian Foundation for Basic
624: Research, grant\#02-02-16664, and the Russian program of
625: scientific school support (\#166.2003.02).
626:
627: % Create the reference section using BibTeX:
628: \begin{thebibliography}{99}
629:
630: \bibitem{NC} see for example M.A. Nielsen and I.L. Chuang, { \it Quantum
631: computation and Information}, Cambridge 2000; D. Bouwmeester et
632: al., { \it The physics of quantum information}, Springer 2000; N.
633: Gisin et al., quant-ph 0101098 and ref.s therein.
634: \bibitem{teleth} C.H. Bennett et al., \PRL {\bf 70}, 1895 (1993).
635:
636: \bibitem{tele} D. Bouwmeester, J. Pan, M. Eibi, H. weinfurter and A. Zeilinger,
637: "Experimental quantum teleportation" , Nature {\bf 390}, 575
638: (1997); D. Boschi, S. Branca, F. De Martini, L. Hardy and S.
639: Popescu, "Experimental realization of teleporting an unknown pure
640: quantum state via dual classical and Einstein-Podolsky-Rosen
641: channels", \PRL {\bf 80} (98) 1121; Y.-H. Kim, S.Kulik and Y.
642: Shih, "Quantum teleportation of a polarization state with a
643: complete Bell state measurement", \PRL {\bf 86}, 1370 (2001).
644:
645: \bibitem{swex} J.-W. Pan et al., \PRL {\bf 80}, 3891 (1998); T.
646: Jennewein, G. Weihs, J. Pan and A. Zeilinger, "Experimental
647: Nonlocality proof of quantum teleportation and entanglement
648: swapping", \PRL {\bf 88}, 017903 (2002).
649:
650: \bibitem{deMartini} S.Giacomini, F. Sciarrino, E. Lombardi and F. De Martini, "Active teleportation of
651: a quantum qubit",Phys. Rev. A {\bf 66}, 030302(R) (2002).
652:
653: \bibitem{pitt1} T.B. Pittman, B.C. Jacobs and J.D. Franson, ''Single
654: photons on pseudodemand from stored parametric down-conversion'',
655: Phys. Rev. A {\bf 66} (2002), 042303;
656:
657: \bibitem{pitt2} T.B. Pittman, B.C. Jacobs,
658: and J.D. Franson, ''Demonstration of feed-forward control for
659: linear optics quantum computation'', Phys. Rev. A {\bf 66} (2002),
660: 052305.
661:
662: \bibitem{au} G. Auletta, {\it Foundations and interpretation of quantum mechanics}
663: (World Scientific, Singapore 2000) and ref.s therein.
664:
665: \bibitem{bell} A. Aspect, J. Dalibard, G. Roger, ''Experimental Test of
666: Bell's Inequalities Using Time- Varying Analyzers'', Phys. Rev. Lett. {\bf 49%
667: } (1982) 1804; J. G. Rarity and P. R. Tapster, ''Experimental
668: violation of
669: Bell's inequality based on phase and momentum'', {Phys. Rev. Lett.} {\bf 64 }%
670: (1990), 2495; P. G. Kwiat et al, ''Correlated two-photon
671: interference in a dual-beam Michelson interferometer'',
672: {Phys.Rev.} A. {\bf 41 }(1990), 2910; W. Tittel, J. Brendel, H.
673: Zbinden, N. Gisin, ''Violation of Bell inequalities by photons
674: more than 10 km apart'', Phys.Rev. Lett. {\bf 81} (1998), 3563; E.
675: Kiess, Y. H. Shih, A. V. Sergienko and C. O. Alley,
676: Einstein-Podolsky-Rosen-Bohm Experiment Using Pairs of Light
677: Quanta Produced by Type-II Parametric Down Conversion'', Phys.
678: Rev. Lett. {\bf 71} (1993), 3893; P. G. Kwiat, K. Mattle, A. V.
679: Sergienko, Y. H. Shih, H. Weinfurter, A. Zeilinger, ''New
680: high-intensity source of polarization-entangled photon pairs'',
681: Phys. Rev. Lett. {\bf 75} (1995), 4337.
682:
683: \bibitem{QCr} See, for example, N. Gisin, G. Ribordy, W. Tittel and H.
684: Zbinden, ''Quantum cryptography'', Rev. Mod. Phys. {\bf 74}
685: (2002), 145-195, and refs. therein.
686:
687: \bibitem{QCo} T.B. Pittman, M. J. Fitch, B. C. Jacobs, and J. D. Franson,
688: ''Experimental Controlled-NOT Logic Gate for Single Photons in the
689: Coincidence Basis et al., quant-ph 0303095; T. B. Pittman, B. C.
690: Jacobs, and J. D. Franson, ''Demonstration of nondeterministic
691: quantum logic operations using linear optical elements'', Phys.
692: Rev. Lett. {\bf 88} (2002), 257902; E. Knill, R. Laflamme, G. J.
693: Milburn, ''A scheme for efficient quantum computation with linear
694: optics'', Nature {\bf 409} (2001), 46.
695:
696:
697: \bibitem{typeI} A. G. White, Daniel F. V. James, Philippe H. Eberhard, and
698: Paul G. Kwiat, ''Nonmaximally Entangled States: Production,
699: Characterization, and Utilization'', Phys. Rev. Lett. {\bf 83}
700: (1999), 3103; Y.Kim, S.P.Kulik, Y.Shih, ''Bell State Preparation
701: Using Pulsed Nondegenerate Two-Photon Entanglement'', Phys. Rev.
702: {\bf 63} (2001), 060301; Y.H.Kim, M.V.Chekhova, S.P.Kulik,
703: M.Rubin, and Y.H.Shih, ''Interferometric Bell State Preparation
704: Using Femtosecond Pulse Pumped Spontaneous Parametric
705: Down-Conversion'', Phys. Rev. {\bf 63} (2001), 062301; G. Brida,
706: M. Genovese, C. Novero and E. Predazzi, ''New experimental test of
707: Bell Inequalities by the use of a Non-Maximally Entangled photon
708: state'', Phys. Lett. A {\bf 268} (2000), 12; G. Brida, M.
709: Genovese, M. Gramegna, C. Novero and E. Predazzi, "A first test of
710: Wigner function local realistic model", Phys. Lett. A {\bf 299}
711: (2002), 121;M. Fiorentino, G. Messin, C. Kuklewicz, F.N.C. Wong
712: and J.H. Shapiro, "Generation of ultrabright tunable polarization
713: entanglement without spatial, spectral or temporal constraints",
714: \PRA {\bf 69} 041801 (R), 2004.
715:
716:
717: \bibitem{bp1} D.C. Burnham and D.L. Weinberg, ''Observation of Simultaneity
718: in Parametric Production of Optical Photon Pairs'', Phys. Rev.
719: Lett. 25 (1970), 84; D.N. Klyshko, Sov. J. Quant. Elect. {\bf 10},
720: 1112 (1980); A. A. Malygin, A. N. Penin, A. V. Sergienko,
721: ''Absolute Calibration of the Sensitivity of Photodetectors Using
722: a Two-Photon Field'', Sov. Phys. JETP Lett. {\bf 33} (1981) 477.
723:
724: \bibitem{alan} A. Migdall, ''Correlated-photon metrology without absolute
725: standards'', Physics Today (1999), 41-46, and refs. therein.
726:
727:
728: \bibitem{bp2} G. Brida, M. Genovese and C. Novero, An
729: application of two photons entangled states to quantum
730: metrology'', Jour. Mod. Opt. {\bf 47} (2000), 2099, and refs.
731: therein.
732:
733: \bibitem{nosCPEM} G. Brida et al., work in progress, to appear in proc. of CPEM
734: 2004, London.
735:
736: \bibitem{pd}D.N.Klyshko, JETP {\bf
737: 84} (6), 1065 (1997).
738: \end{thebibliography}
739:
740: \newpage
741: {\bf Figure Captions} \vskip 2cm
742: Fig.1 Our set-up. A cw argon
743: laser generating at 351 nm pumps a type-II BBO crystal cut for
744: frequency-degenerate non-collinear polarization-entangled phase
745: matching. A quartz crystal compensate birefringence. One of the
746: correlated photons, after a spatial selection by means of a
747: pinhole A, a spectral selection by means of a red-glass cutoff
748: filter RG, and a polarization selection by means of a polarizing
749: cube PBS, is addressed to the photon counter D1, which drives,
750: through a fast high-voltage switch S, a Pockels cell PC placed in
751: the optical path of the other photon. The delay between a
752: photocount of D1 and the corresponding high-voltage pulse on the
753: Pockels cell can be varied electronically. The second photon of
754: the entangled pair is retarded, before the Pockels cell, by means
755: of fiber F. This realizes the conditioned unitary operation. The
756: second photon is registered by photon counter D2 preceded by a
757: Glan prism G and an interference filter IF. The output signals
758: from the detectors are routed to a two channel counter C, in
759: order to have the number of events on single channel, and to a
760: Time to Amplitude Converter circuit, followed by a single channel
761: analyzer, for selecting and counting coincidence events.
762:
763: Fig. 2 Picture of our optical bench. One can recognize the two
764: detectors preceded by
765: filters,iris and polarizers,
766: the Pockels cell and the fiber.
767:
768: Fig. 3 Counting rate of the second detector as a function of the
769: angle of the polarizer preceding it (without background
770: subtraction) in absence of the Pockels cell.
771:
772: Fig. 4 Coincidences as a function of the angle of the polarizer
773: preceding the second detector for the Bell state $\vert \phi ^-
774: \rangle = {\frac{ \vert 45 \rangle \vert 45 \rangle - \vert 135
775: \rangle \vert 135 \rangle }{\sqrt {2}}} \label{phi}$. When the
776: Pockels cell (KDP) is not activated the maximum is at $135^o$
777: (squares). When a $90^o$ rotation of polarization is realized by
778: the Pockels cell conditioned to the measurement of a $45 ^o$
779: polarized photon in the conjugated arm, the maximum is shifted, as
780: expected, to $45^o$ (triangles).
781:
782: Fig. 5 Von Neumann entropy of the trigger photon polarization
783: density matrix
784: as a function of the quantum efficiency $\eta_1$ of the first detector. When $\eta_1=1$ the final
785: polarization reduced density matrix $\rho _2$ of the second photon
786: corresponds to a pure state, giving $S(\rho_2)=0$. On the other
787: hand if $\eta_1=0$ the final polarization reduced density matrix
788: of the second photon is completely mixed with $S(\rho _2)=1$.
789: \label{S}
790:
791: Fig. 6 Counts on the second detector as a function of the angle of
792: the polarizer preceding it (without background subtraction). When
793: the Pockels cell (KDP) is not activated no dependence on the
794: polarizer angle appears (squares). When a $90^o$ rotation of
795: polarization is realized by the Pockels cell conditioned to the
796: measurement of a vertically polarized photon on the conjugated
797: branch, the data (triangles) show a clear dependence on the
798: polarizer setting (corresponding to a mainly vertically polarized
799: state).
800:
801: Fig. 7 Coincidences as a function of the angle of the polarizer
802: preceding the second detector for the state described by the
803: density matrix $\rho =1/2 ( \eta_1 \vert V \rangle \langle V \vert
804: + (1- \eta_1) \vert H \rangle \langle H \vert)$ . When the Pockels
805: cell (KDP) is not activated a maximum is at $0^o$ (squares). When
806: a $90^o$ rotation of polarization is realized by the Pockels cell
807: conditioned to the measurement of a vertically polarized photon on
808: the conjugated branch, the maximum is shifted, as expected, by
809: $90^o$ (triangles).
810:
811: Fig. 8 The pulse generated by the Pockels cell drive and the
812: light measured after the Pockels cell between crossed polarizers
813: when activated by the former pulse and reached by a He-Ne laser
814: beam.
815:
816: Fig. 9 Dependence of the D2 counting rate at $\theta=0^o$
817: (Horizontally polarized photons, triangles) and $\theta=90^o$
818: (Vertically polarized photons, squares) on the delay $T$
819: introduced electronically between the trigger pulses from D1
820: detector and the corresponding high-voltage pulses driving the
821: Pockel's cell. The effect of counting rate decreasing for
822: horizontally polarized photons is observed for $T < 100 \, ns$
823: and not observed for $T>100 \, ns$.
824:
825: Fig. 10 Dependence of coincidences at $\theta=0^o$ (Horizontally
826: polarized photons, triangles) and $\theta=90^o$ (Vertically
827: polarized photons, squares) for the polarization selection of the
828: second photon on the delay $T$ introduced electronically between
829: the trigger pulses from D1 detector and the corresponding
830: high-voltage pulses driving the Pockel's cell. The effect of
831: reversing of coincidences between H and V selection is observed
832: at $T < 100 \, ns$.
833:
834: Fig. 11 Coincidences in function of the counts on the detector
835: under calibration (by varying the power of pump laser) both with
836: (circles) and without (squares) corrections for detector dead time
837: and for stop delay time \cite{bp2}. These data are used for
838: evaluating the quantum efficiency of the signal detector by means
839: of traditional scheme for biphoton calibration of detectors
840: \cite{bp1,alan}.
841:
842:
843:
844: \end{document}
845: